The C based gRPC (C++, Python, Ruby, Objective-C, PHP, C#) https://grpc.io/
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

2342 lines
81 KiB

//
//
// Copyright 2015 gRPC authors.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
//
//
10 years ago
#include <mutex>
#include <thread>
#include "absl/log/check.h"
#include "absl/log/log.h"
#include "absl/memory/memory.h"
#include "absl/strings/ascii.h"
#include "absl/strings/match.h"
#include "absl/strings/str_format.h"
9 years ago
#include <grpc/grpc.h>
7 years ago
#include <grpc/support/alloc.h>
9 years ago
#include <grpc/support/time.h>
#include <grpcpp/channel.h>
#include <grpcpp/client_context.h>
#include <grpcpp/create_channel.h>
#include <grpcpp/resource_quota.h>
#include <grpcpp/security/auth_metadata_processor.h>
#include <grpcpp/security/credentials.h>
#include <grpcpp/security/server_credentials.h>
#include <grpcpp/server.h>
#include <grpcpp/server_builder.h>
#include <grpcpp/server_context.h>
#include <grpcpp/support/string_ref.h>
#include <grpcpp/test/channel_test_peer.h>
10 years ago
#include "src/core/client_channel/backup_poller.h"
#include "src/core/lib/config/config_vars.h"
#include "src/core/lib/gprpp/crash.h"
#include "src/core/lib/gprpp/env.h"
#include "src/core/lib/iomgr/iomgr.h"
#include "src/core/lib/security/credentials/credentials.h"
9 years ago
#include "src/proto/grpc/testing/duplicate/echo_duplicate.grpc.pb.h"
#include "src/proto/grpc/testing/echo.grpc.pb.h"
#include "test/core/test_util/port.h"
#include "test/core/test_util/test_config.h"
#include "test/cpp/end2end/interceptors_util.h"
#include "test/cpp/end2end/test_service_impl.h"
9 years ago
#include "test/cpp/util/string_ref_helper.h"
#include "test/cpp/util/test_credentials_provider.h"
10 years ago
#ifdef GRPC_POSIX_SOCKET_EV
#include "src/core/lib/iomgr/ev_posix.h"
#endif // GRPC_POSIX_SOCKET_EV
#include <gtest/gtest.h>
using std::chrono::system_clock;
10 years ago
namespace grpc {
namespace testing {
namespace {
bool CheckIsLocalhost(const std::string& addr) {
const std::string kIpv6("ipv6:%5B::1%5D:");
const std::string kIpv4MappedIpv6("ipv6:%5B::ffff:127.0.0.1%5D:");
const std::string kIpv4("ipv4:127.0.0.1:");
10 years ago
return addr.substr(0, kIpv4.size()) == kIpv4 ||
addr.substr(0, kIpv4MappedIpv6.size()) == kIpv4MappedIpv6 ||
addr.substr(0, kIpv6.size()) == kIpv6;
}
const int kClientChannelBackupPollIntervalMs = 200;
const char kTestCredsPluginErrorMsg[] = "Could not find plugin metadata.";
const char kFakeToken[] = "fake_token";
const char kFakeSelector[] = "fake_selector";
const char kExpectedFakeCredsDebugString[] =
"CallCredentials{GoogleIAMCredentials{Token:present,"
"AuthoritySelector:fake_selector}}";
const char kWrongToken[] = "wrong_token";
const char kWrongSelector[] = "wrong_selector";
const char kExpectedWrongCredsDebugString[] =
"CallCredentials{GoogleIAMCredentials{Token:present,"
"AuthoritySelector:wrong_selector}}";
const char kFakeToken1[] = "fake_token1";
const char kFakeSelector1[] = "fake_selector1";
const char kExpectedFakeCreds1DebugString[] =
"CallCredentials{GoogleIAMCredentials{Token:present,"
"AuthoritySelector:fake_selector1}}";
const char kFakeToken2[] = "fake_token2";
const char kFakeSelector2[] = "fake_selector2";
const char kExpectedFakeCreds2DebugString[] =
"CallCredentials{GoogleIAMCredentials{Token:present,"
"AuthoritySelector:fake_selector2}}";
const char kExpectedAuthMetadataPluginKeyFailureCredsDebugString[] =
"CallCredentials{TestMetadataCredentials{key:TestPluginMetadata,"
"value:Does not matter, will fail the key is invalid.}}";
const char kExpectedAuthMetadataPluginValueFailureCredsDebugString[] =
"CallCredentials{TestMetadataCredentials{key:test-plugin-metadata,"
"value:With illegal \n value.}}";
const char kExpectedAuthMetadataPluginWithDeadlineCredsDebugString[] =
"CallCredentials{TestMetadataCredentials{key:meta_key,value:Does "
"not "
"matter}}";
const char kExpectedNonBlockingAuthMetadataPluginFailureCredsDebugString[] =
"CallCredentials{TestMetadataCredentials{key:test-plugin-metadata,"
"value:Does not matter, will fail anyway (see 3rd param)}}";
const char
kExpectedNonBlockingAuthMetadataPluginAndProcessorSuccessCredsDebugString
[] = "CallCredentials{TestMetadataCredentials{key:test-plugin-"
"metadata,value:Dr Jekyll}}";
const char
kExpectedNonBlockingAuthMetadataPluginAndProcessorFailureCredsDebugString
[] = "CallCredentials{TestMetadataCredentials{key:test-plugin-"
"metadata,value:Mr Hyde}}";
const char kExpectedBlockingAuthMetadataPluginFailureCredsDebugString[] =
"CallCredentials{TestMetadataCredentials{key:test-plugin-metadata,"
"value:Does not matter, will fail anyway (see 3rd param)}}";
const char kExpectedCompositeCallCredsDebugString[] =
"CallCredentials{CompositeCallCredentials{TestMetadataCredentials{"
"key:call-creds-key1,value:call-creds-val1},TestMetadataCredentials{key:"
"call-creds-key2,value:call-creds-val2}}}";
class TestMetadataCredentialsPlugin : public MetadataCredentialsPlugin {
public:
static const char kGoodMetadataKey[];
static const char kBadMetadataKey[];
TestMetadataCredentialsPlugin(const grpc::string_ref& metadata_key,
const grpc::string_ref& metadata_value,
bool is_blocking, bool is_successful,
int delay_ms)
: metadata_key_(metadata_key.data(), metadata_key.length()),
metadata_value_(metadata_value.data(), metadata_value.length()),
is_blocking_(is_blocking),
is_successful_(is_successful),
delay_ms_(delay_ms) {}
bool IsBlocking() const override { return is_blocking_; }
8 years ago
Status GetMetadata(
grpc::string_ref service_url, grpc::string_ref method_name,
const grpc::AuthContext& channel_auth_context,
std::multimap<std::string, std::string>* metadata) override {
if (delay_ms_ != 0) {
gpr_sleep_until(
gpr_time_add(gpr_now(GPR_CLOCK_REALTIME),
gpr_time_from_millis(delay_ms_, GPR_TIMESPAN)));
}
EXPECT_GT(service_url.length(), 0UL);
EXPECT_GT(method_name.length(), 0UL);
EXPECT_TRUE(channel_auth_context.IsPeerAuthenticated());
EXPECT_TRUE(metadata != nullptr);
if (is_successful_) {
metadata->insert(std::make_pair(metadata_key_, metadata_value_));
return Status::OK;
} else {
return Status(StatusCode::NOT_FOUND, kTestCredsPluginErrorMsg);
}
}
std::string DebugString() override {
return absl::StrFormat("TestMetadataCredentials{key:%s,value:%s}",
metadata_key_.c_str(), metadata_value_.c_str());
}
private:
std::string metadata_key_;
std::string metadata_value_;
bool is_blocking_;
bool is_successful_;
int delay_ms_;
};
const char TestMetadataCredentialsPlugin::kBadMetadataKey[] =
"TestPluginMetadata";
const char TestMetadataCredentialsPlugin::kGoodMetadataKey[] =
"test-plugin-metadata";
class TestAuthMetadataProcessor : public AuthMetadataProcessor {
public:
static const char kGoodGuy[];
explicit TestAuthMetadataProcessor(bool is_blocking)
: is_blocking_(is_blocking) {}
std::shared_ptr<CallCredentials> GetCompatibleClientCreds() {
return grpc::MetadataCredentialsFromPlugin(
std::unique_ptr<MetadataCredentialsPlugin>(
new TestMetadataCredentialsPlugin(
TestMetadataCredentialsPlugin::kGoodMetadataKey, kGoodGuy,
is_blocking_, true, 0)));
}
std::shared_ptr<CallCredentials> GetIncompatibleClientCreds() {
return grpc::MetadataCredentialsFromPlugin(
std::unique_ptr<MetadataCredentialsPlugin>(
new TestMetadataCredentialsPlugin(
TestMetadataCredentialsPlugin::kGoodMetadataKey, "Mr Hyde",
is_blocking_, true, 0)));
}
// Interface implementation
bool IsBlocking() const override { return is_blocking_; }
Status Process(const InputMetadata& auth_metadata, AuthContext* context,
OutputMetadata* consumed_auth_metadata,
OutputMetadata* response_metadata) override {
EXPECT_TRUE(consumed_auth_metadata != nullptr);
EXPECT_TRUE(context != nullptr);
EXPECT_TRUE(response_metadata != nullptr);
auto auth_md =
auth_metadata.find(TestMetadataCredentialsPlugin::kGoodMetadataKey);
EXPECT_NE(auth_md, auth_metadata.end());
string_ref auth_md_value = auth_md->second;
if (auth_md_value == kGoodGuy) {
context->AddProperty(kIdentityPropName, kGoodGuy);
context->SetPeerIdentityPropertyName(kIdentityPropName);
consumed_auth_metadata->insert(std::make_pair(
string(auth_md->first.data(), auth_md->first.length()),
string(auth_md->second.data(), auth_md->second.length())));
return Status::OK;
} else {
return Status(StatusCode::UNAUTHENTICATED,
string("Invalid principal: ") +
string(auth_md_value.data(), auth_md_value.length()));
}
}
private:
static const char kIdentityPropName[];
bool is_blocking_;
};
const char TestAuthMetadataProcessor::kGoodGuy[] = "Dr Jekyll";
const char TestAuthMetadataProcessor::kIdentityPropName[] = "novel identity";
class Proxy : public grpc::testing::EchoTestService::Service {
public:
explicit Proxy(const std::shared_ptr<Channel>& channel)
: stub_(grpc::testing::EchoTestService::NewStub(channel)) {}
Status Echo(ServerContext* server_context, const EchoRequest* request,
EchoResponse* response) override {
std::unique_ptr<ClientContext> client_context =
ClientContext::FromServerContext(*server_context);
return stub_->Echo(client_context.get(), *request, response);
}
private:
std::unique_ptr<grpc::testing::EchoTestService::Stub> stub_;
};
class TestServiceImplDupPkg
: public grpc::testing::duplicate::EchoTestService::Service {
public:
Status Echo(ServerContext* /*context*/, const EchoRequest* /*request*/,
EchoResponse* response) override {
response->set_message("no package");
return Status::OK;
}
};
class TestScenario {
public:
TestScenario(bool use_interceptors, bool use_proxy, bool inproc,
const std::string& credentials_type, bool callback_server)
: use_interceptors_(use_interceptors),
use_proxy_(use_proxy),
inproc_(inproc),
credentials_type_(credentials_type),
callback_server_(callback_server) {}
bool use_interceptors() const { return use_interceptors_; }
bool use_proxy() const { return use_proxy_; }
bool inproc() const { return inproc_; }
const std::string& credentials_type() const { return credentials_type_; }
bool callback_server() const { return callback_server_; }
std::string AsString() const;
static std::string Name(const ::testing::TestParamInfo<TestScenario>& info) {
return info.param.AsString();
}
private:
bool use_interceptors_;
bool use_proxy_;
bool inproc_;
const std::string credentials_type_;
bool callback_server_;
};
std::string TestScenario::AsString() const {
std::string retval = use_interceptors_ ? "Interceptor" : "";
if (use_proxy_) retval += "Proxy";
if (inproc_) retval += "Inproc";
if (callback_server_) retval += "CallbackServer";
if (credentials_type_ == kInsecureCredentialsType) {
retval += "Insecure";
} else {
std::string creds_type = absl::AsciiStrToLower(credentials_type_);
if (!creds_type.empty()) creds_type[0] = absl::ascii_toupper(creds_type[0]);
retval += creds_type;
}
return retval;
}
class End2endTest : public ::testing::TestWithParam<TestScenario> {
10 years ago
protected:
static void SetUpTestSuite() { grpc_init(); }
static void TearDownTestSuite() { grpc_shutdown(); }
End2endTest()
: is_server_started_(false),
kMaxMessageSize_(8192),
special_service_("special"),
first_picked_port_(0) {}
void TearDown() override {
if (is_server_started_) {
server_->Shutdown();
if (proxy_server_) proxy_server_->Shutdown();
}
if (first_picked_port_ > 0) {
grpc_recycle_unused_port(first_picked_port_);
}
}
void StartServer(const std::shared_ptr<AuthMetadataProcessor>& processor) {
int port = grpc_pick_unused_port_or_die();
first_picked_port_ = port;
server_address_ << "localhost:" << port;
10 years ago
// Setup server
BuildAndStartServer(processor);
}
void RestartServer(const std::shared_ptr<AuthMetadataProcessor>& processor) {
if (is_server_started_) {
server_->Shutdown();
BuildAndStartServer(processor);
}
}
void BuildAndStartServer(
const std::shared_ptr<AuthMetadataProcessor>& processor) {
10 years ago
ServerBuilder builder;
ConfigureServerBuilder(&builder);
auto server_creds = GetCredentialsProvider()->GetServerCredentials(
GetParam().credentials_type());
if (GetParam().credentials_type() != kInsecureCredentialsType) {
server_creds->SetAuthMetadataProcessor(processor);
}
if (GetParam().use_interceptors()) {
std::vector<
std::unique_ptr<experimental::ServerInterceptorFactoryInterface>>
creators;
// Add 20 phony server interceptors
creators.reserve(20);
for (auto i = 0; i < 20; i++) {
creators.push_back(std::make_unique<PhonyInterceptorFactory>());
}
builder.experimental().SetInterceptorCreators(std::move(creators));
}
builder.AddListeningPort(server_address_.str(), server_creds);
if (!GetParam().callback_server()) {
builder.RegisterService(&service_);
} else {
builder.RegisterService(&callback_service_);
}
builder.RegisterService("foo.test.youtube.com", &special_service_);
builder.RegisterService(&dup_pkg_service_);
8 years ago
builder.SetSyncServerOption(ServerBuilder::SyncServerOption::NUM_CQS, 4);
builder.SetSyncServerOption(
ServerBuilder::SyncServerOption::CQ_TIMEOUT_MSEC, 10);
10 years ago
server_ = builder.BuildAndStart();
is_server_started_ = true;
}
10 years ago
virtual void ConfigureServerBuilder(ServerBuilder* builder) {
builder->SetMaxMessageSize(
kMaxMessageSize_); // For testing max message size.
}
void ResetChannel(
std::vector<
std::unique_ptr<experimental::ClientInterceptorFactoryInterface>>
interceptor_creators = {}) {
if (!is_server_started_) {
StartServer(std::shared_ptr<AuthMetadataProcessor>());
}
EXPECT_TRUE(is_server_started_);
ChannelArguments args;
auto channel_creds = GetCredentialsProvider()->GetChannelCredentials(
GetParam().credentials_type(), &args);
if (!user_agent_prefix_.empty()) {
args.SetUserAgentPrefix(user_agent_prefix_);
}
args.SetString(GRPC_ARG_SECONDARY_USER_AGENT_STRING, "end2end_test");
if (!GetParam().inproc()) {
if (!GetParam().use_interceptors()) {
channel_ = grpc::CreateCustomChannel(server_address_.str(),
channel_creds, args);
} else {
channel_ = CreateCustomChannelWithInterceptors(
server_address_.str(), channel_creds, args,
interceptor_creators.empty() ? CreatePhonyClientInterceptors()
: std::move(interceptor_creators));
}
} else {
if (!GetParam().use_interceptors()) {
channel_ = server_->InProcessChannel(args);
} else {
channel_ = server_->experimental().InProcessChannelWithInterceptors(
args, interceptor_creators.empty()
? CreatePhonyClientInterceptors()
: std::move(interceptor_creators));
}
}
}
void ResetStub(
std::vector<
std::unique_ptr<experimental::ClientInterceptorFactoryInterface>>
interceptor_creators = {}) {
ResetChannel(std::move(interceptor_creators));
if (GetParam().use_proxy()) {
proxy_service_ = std::make_unique<Proxy>(channel_);
int port = grpc_pick_unused_port_or_die();
std::ostringstream proxyaddr;
proxyaddr << "localhost:" << port;
ServerBuilder builder;
builder.AddListeningPort(proxyaddr.str(), InsecureServerCredentials());
builder.RegisterService(proxy_service_.get());
8 years ago
builder.SetSyncServerOption(ServerBuilder::SyncServerOption::NUM_CQS, 4);
builder.SetSyncServerOption(
ServerBuilder::SyncServerOption::CQ_TIMEOUT_MSEC, 10);
proxy_server_ = builder.BuildAndStart();
channel_ =
grpc::CreateChannel(proxyaddr.str(), InsecureChannelCredentials());
}
stub_ = grpc::testing::EchoTestService::NewStub(channel_);
PhonyInterceptor::Reset();
}
bool is_server_started_;
std::shared_ptr<Channel> channel_;
std::unique_ptr<grpc::testing::EchoTestService::Stub> stub_;
10 years ago
std::unique_ptr<Server> server_;
std::unique_ptr<Server> proxy_server_;
std::unique_ptr<Proxy> proxy_service_;
10 years ago
std::ostringstream server_address_;
const int kMaxMessageSize_;
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
TestServiceImpl service_;
CallbackTestServiceImpl callback_service_;
TestServiceImpl special_service_;
TestServiceImplDupPkg dup_pkg_service_;
std::string user_agent_prefix_;
int first_picked_port_;
10 years ago
};
void SendRpc(grpc::testing::EchoTestService::Stub* stub, int num_rpcs,
bool with_binary_metadata) {
10 years ago
EchoRequest request;
EchoResponse response;
request.set_message("Hello hello hello hello");
10 years ago
for (int i = 0; i < num_rpcs; ++i) {
ClientContext context;
if (with_binary_metadata) {
7 years ago
char bytes[8] = {'\0', '\1', '\2', '\3',
'\4', '\5', '\6', static_cast<char>(i)};
context.AddMetadata("custom-bin", std::string(bytes, 8));
}
context.set_compression_algorithm(GRPC_COMPRESS_GZIP);
10 years ago
Status s = stub->Echo(&context, request, &response);
EXPECT_EQ(response.message(), request.message());
EXPECT_TRUE(s.ok());
10 years ago
}
}
// This class is for testing scenarios where RPCs are cancelled on the server
// by calling ServerContext::TryCancel()
class End2endServerTryCancelTest : public End2endTest {
protected:
// Helper for testing client-streaming RPCs which are cancelled on the server.
// Depending on the value of server_try_cancel parameter, this will test one
// of the following three scenarios:
// CANCEL_BEFORE_PROCESSING: Rpc is cancelled by the server before reading
// any messages from the client
//
// CANCEL_DURING_PROCESSING: Rpc is cancelled by the server while reading
// messages from the client
//
// CANCEL_AFTER PROCESSING: Rpc is cancelled by server after reading all
// the messages from the client
//
// NOTE: Do not call this function with server_try_cancel == DO_NOT_CANCEL.
void TestRequestStreamServerCancel(
ServerTryCancelRequestPhase server_try_cancel, int num_msgs_to_send) {
RestartServer(std::shared_ptr<AuthMetadataProcessor>());
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
// Send server_try_cancel value in the client metadata
context.AddMetadata(kServerTryCancelRequest,
std::to_string(server_try_cancel));
auto stream = stub_->RequestStream(&context, &response);
int num_msgs_sent = 0;
while (num_msgs_sent < num_msgs_to_send) {
request.set_message("hello");
if (!stream->Write(request)) {
break;
}
num_msgs_sent++;
}
LOG(INFO) << "Sent " << num_msgs_sent << " messages";
stream->WritesDone();
Status s = stream->Finish();
// At this point, we know for sure that RPC was cancelled by the server
// since we passed server_try_cancel value in the metadata. Depending on the
// value of server_try_cancel, the RPC might have been cancelled by the
// server at different stages. The following validates our expectations of
// number of messages sent in various cancellation scenarios:
switch (server_try_cancel) {
case CANCEL_BEFORE_PROCESSING:
case CANCEL_DURING_PROCESSING:
// If the RPC is cancelled by server before / during messages from the
// client, it means that the client most likely did not get a chance to
// send all the messages it wanted to send. i.e num_msgs_sent <=
// num_msgs_to_send
EXPECT_LE(num_msgs_sent, num_msgs_to_send);
break;
case CANCEL_AFTER_PROCESSING:
// If the RPC was cancelled after all messages were read by the server,
// the client did get a chance to send all its messages
EXPECT_EQ(num_msgs_sent, num_msgs_to_send);
break;
default:
LOG(ERROR) << "Invalid server_try_cancel value: " << server_try_cancel;
EXPECT_TRUE(server_try_cancel > DO_NOT_CANCEL &&
server_try_cancel <= CANCEL_AFTER_PROCESSING);
break;
}
EXPECT_FALSE(s.ok());
EXPECT_EQ(grpc::StatusCode::CANCELLED, s.error_code());
// Make sure that the server interceptors were notified
if (GetParam().use_interceptors()) {
EXPECT_EQ(20, PhonyInterceptor::GetNumTimesCancel());
}
}
// Helper for testing server-streaming RPCs which are cancelled on the server.
// Depending on the value of server_try_cancel parameter, this will test one
// of the following three scenarios:
// CANCEL_BEFORE_PROCESSING: Rpc is cancelled by the server before writing
// any messages to the client
//
// CANCEL_DURING_PROCESSING: Rpc is cancelled by the server while writing
// messages to the client
//
// CANCEL_AFTER PROCESSING: Rpc is cancelled by server after writing all
// the messages to the client
//
// NOTE: Do not call this function with server_try_cancel == DO_NOT_CANCEL.
void TestResponseStreamServerCancel(
ServerTryCancelRequestPhase server_try_cancel) {
RestartServer(std::shared_ptr<AuthMetadataProcessor>());
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
// Send server_try_cancel in the client metadata
context.AddMetadata(kServerTryCancelRequest,
std::to_string(server_try_cancel));
request.set_message("hello");
auto stream = stub_->ResponseStream(&context, request);
int num_msgs_read = 0;
while (num_msgs_read < kServerDefaultResponseStreamsToSend) {
if (!stream->Read(&response)) {
break;
}
EXPECT_EQ(response.message(),
request.message() + std::to_string(num_msgs_read));
num_msgs_read++;
}
LOG(INFO) << "Read " << num_msgs_read << " messages";
Status s = stream->Finish();
// Depending on the value of server_try_cancel, the RPC might have been
// cancelled by the server at different stages. The following validates our
// expectations of number of messages read in various cancellation
// scenarios:
switch (server_try_cancel) {
case CANCEL_BEFORE_PROCESSING:
// Server cancelled before sending any messages. Which means the client
// wouldn't have read any
EXPECT_EQ(num_msgs_read, 0);
break;
case CANCEL_DURING_PROCESSING:
// Server cancelled while writing messages. Client must have read less
// than or equal to the expected number of messages
EXPECT_LE(num_msgs_read, kServerDefaultResponseStreamsToSend);
break;
case CANCEL_AFTER_PROCESSING:
// Even though the Server cancelled after writing all messages, the RPC
// may be cancelled before the Client got a chance to read all the
// messages.
EXPECT_LE(num_msgs_read, kServerDefaultResponseStreamsToSend);
break;
default: {
LOG(ERROR) << "Invalid server_try_cancel value: " << server_try_cancel;
EXPECT_TRUE(server_try_cancel > DO_NOT_CANCEL &&
server_try_cancel <= CANCEL_AFTER_PROCESSING);
break;
}
}
EXPECT_FALSE(s.ok());
// Make sure that the server interceptors were notified
if (GetParam().use_interceptors()) {
EXPECT_EQ(20, PhonyInterceptor::GetNumTimesCancel());
}
}
// Helper for testing bidirectional-streaming RPCs which are cancelled on the
// server. Depending on the value of server_try_cancel parameter, this will
// test one of the following three scenarios:
// CANCEL_BEFORE_PROCESSING: Rpc is cancelled by the server before reading/
// writing any messages from/to the client
//
// CANCEL_DURING_PROCESSING: Rpc is cancelled by the server while reading/
// writing messages from/to the client
//
// CANCEL_AFTER PROCESSING: Rpc is cancelled by server after reading/writing
// all the messages from/to the client
//
// NOTE: Do not call this function with server_try_cancel == DO_NOT_CANCEL.
void TestBidiStreamServerCancel(ServerTryCancelRequestPhase server_try_cancel,
int num_messages) {
RestartServer(std::shared_ptr<AuthMetadataProcessor>());
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
// Send server_try_cancel in the client metadata
context.AddMetadata(kServerTryCancelRequest,
std::to_string(server_try_cancel));
auto stream = stub_->BidiStream(&context);
int num_msgs_read = 0;
int num_msgs_sent = 0;
while (num_msgs_sent < num_messages) {
request.set_message("hello " + std::to_string(num_msgs_sent));
if (!stream->Write(request)) {
break;
}
num_msgs_sent++;
if (!stream->Read(&response)) {
break;
}
num_msgs_read++;
EXPECT_EQ(response.message(), request.message());
}
LOG(INFO) << "Sent " << num_msgs_sent << " messages";
LOG(INFO) << "Read " << num_msgs_read << " messages";
stream->WritesDone();
Status s = stream->Finish();
// Depending on the value of server_try_cancel, the RPC might have been
// cancelled by the server at different stages. The following validates our
// expectations of number of messages read in various cancellation
// scenarios:
switch (server_try_cancel) {
case CANCEL_BEFORE_PROCESSING:
EXPECT_EQ(num_msgs_read, 0);
break;
case CANCEL_DURING_PROCESSING:
EXPECT_LE(num_msgs_sent, num_messages);
EXPECT_LE(num_msgs_read, num_msgs_sent);
break;
case CANCEL_AFTER_PROCESSING:
EXPECT_EQ(num_msgs_sent, num_messages);
// The Server cancelled after reading the last message and after writing
// the message to the client. However, the RPC cancellation might have
// taken effect before the client actually read the response.
EXPECT_LE(num_msgs_read, num_msgs_sent);
break;
default:
LOG(ERROR) << "Invalid server_try_cancel value: " << server_try_cancel;
EXPECT_TRUE(server_try_cancel > DO_NOT_CANCEL &&
server_try_cancel <= CANCEL_AFTER_PROCESSING);
break;
}
EXPECT_FALSE(s.ok());
EXPECT_EQ(grpc::StatusCode::CANCELLED, s.error_code());
// Make sure that the server interceptors were notified
if (GetParam().use_interceptors()) {
EXPECT_EQ(20, PhonyInterceptor::GetNumTimesCancel());
}
}
};
TEST_P(End2endServerTryCancelTest, RequestEchoServerCancel) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
context.AddMetadata(kServerTryCancelRequest,
std::to_string(CANCEL_BEFORE_PROCESSING));
Status s = stub_->Echo(&context, request, &response);
EXPECT_FALSE(s.ok());
EXPECT_EQ(grpc::StatusCode::CANCELLED, s.error_code());
}
// Server to cancel before doing reading the request
TEST_P(End2endServerTryCancelTest, RequestStreamServerCancelBeforeReads) {
TestRequestStreamServerCancel(CANCEL_BEFORE_PROCESSING, 1);
}
// Server to cancel while reading a request from the stream in parallel
TEST_P(End2endServerTryCancelTest, RequestStreamServerCancelDuringRead) {
TestRequestStreamServerCancel(CANCEL_DURING_PROCESSING, 10);
}
// Server to cancel after reading all the requests but before returning to the
// client
TEST_P(End2endServerTryCancelTest, RequestStreamServerCancelAfterReads) {
TestRequestStreamServerCancel(CANCEL_AFTER_PROCESSING, 4);
}
// Server to cancel before sending any response messages
TEST_P(End2endServerTryCancelTest, ResponseStreamServerCancelBefore) {
TestResponseStreamServerCancel(CANCEL_BEFORE_PROCESSING);
}
// Server to cancel while writing a response to the stream in parallel
TEST_P(End2endServerTryCancelTest, ResponseStreamServerCancelDuring) {
TestResponseStreamServerCancel(CANCEL_DURING_PROCESSING);
}
// Server to cancel after writing all the respones to the stream but before
// returning to the client
TEST_P(End2endServerTryCancelTest, ResponseStreamServerCancelAfter) {
TestResponseStreamServerCancel(CANCEL_AFTER_PROCESSING);
}
// Server to cancel before reading/writing any requests/responses on the stream
TEST_P(End2endServerTryCancelTest, BidiStreamServerCancelBefore) {
TestBidiStreamServerCancel(CANCEL_BEFORE_PROCESSING, 2);
}
// Server to cancel while reading/writing requests/responses on the stream in
// parallel
TEST_P(End2endServerTryCancelTest, BidiStreamServerCancelDuring) {
TestBidiStreamServerCancel(CANCEL_DURING_PROCESSING, 10);
}
// Server to cancel after reading/writing all requests/responses on the stream
// but before returning to the client
TEST_P(End2endServerTryCancelTest, BidiStreamServerCancelAfter) {
TestBidiStreamServerCancel(CANCEL_AFTER_PROCESSING, 5);
}
TEST_P(End2endTest, SimpleRpcWithCustomUserAgentPrefix) {
// User-Agent is an HTTP header for HTTP transports only
if (GetParam().inproc()) {
return;
}
user_agent_prefix_ = "custom_prefix";
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("Hello hello hello hello");
request.mutable_param()->set_echo_metadata(true);
ClientContext context;
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(response.message(), request.message());
EXPECT_TRUE(s.ok());
const auto& trailing_metadata = context.GetServerTrailingMetadata();
auto iter = trailing_metadata.find("user-agent");
EXPECT_TRUE(iter != trailing_metadata.end());
std::string expected_prefix = user_agent_prefix_ + " grpc-c++/";
EXPECT_TRUE(iter->second.starts_with(expected_prefix)) << iter->second;
}
TEST_P(End2endTest, MultipleRpcsWithVariedBinaryMetadataValue) {
ResetStub();
std::vector<std::thread> threads;
threads.reserve(10);
for (int i = 0; i < 10; ++i) {
threads.emplace_back(SendRpc, stub_.get(), 10, true);
}
for (int i = 0; i < 10; ++i) {
threads[i].join();
}
}
TEST_P(End2endTest, MultipleRpcs) {
ResetStub();
std::vector<std::thread> threads;
threads.reserve(10);
for (int i = 0; i < 10; ++i) {
threads.emplace_back(SendRpc, stub_.get(), 10, false);
}
for (int i = 0; i < 10; ++i) {
threads[i].join();
}
}
TEST_P(End2endTest, ManyStubs) {
ResetStub();
ChannelTestPeer peer(channel_.get());
int registered_calls_pre = peer.registered_calls();
for (int i = 0; i < 1000; ++i) {
grpc::testing::EchoTestService::NewStub(channel_);
}
EXPECT_EQ(peer.registered_calls(), registered_calls_pre);
}
TEST_P(End2endTest, EmptyBinaryMetadata) {
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("Hello hello hello hello");
ClientContext context;
context.AddMetadata("custom-bin", "");
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(response.message(), request.message());
EXPECT_TRUE(s.ok());
}
TEST_P(End2endTest, AuthoritySeenOnServerSide) {
ResetStub();
EchoRequest request;
request.mutable_param()->set_echo_host_from_authority_header(true);
EchoResponse response;
request.set_message("Live long and prosper.");
ClientContext context;
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(response.message(), request.message());
if (GetParam().credentials_type() == kTlsCredentialsType) {
// SSL creds overrides the authority.
EXPECT_EQ("foo.test.google.fr", response.param().host());
} else if (GetParam().inproc()) {
EXPECT_EQ("inproc", response.param().host());
} else {
EXPECT_EQ(server_address_.str(), response.param().host());
}
EXPECT_TRUE(s.ok());
}
TEST_P(End2endTest, ReconnectChannel) {
if (GetParam().inproc()) {
return;
}
7 years ago
int poller_slowdown_factor = 1;
// It needs 2 pollset_works to reconnect the channel with polling engine
// "poll"
#ifdef GRPC_POSIX_SOCKET_EV
if (grpc_core::ConfigVars::Get().PollStrategy() == "poll") {
7 years ago
poller_slowdown_factor = 2;
}
#endif // GRPC_POSIX_SOCKET_EV
ResetStub();
SendRpc(stub_.get(), 1, false);
RestartServer(std::shared_ptr<AuthMetadataProcessor>());
// It needs more than GRPC_CLIENT_CHANNEL_BACKUP_POLL_INTERVAL_MS time to
// reconnect the channel. Make it a factor of 5x
gpr_sleep_until(
gpr_time_add(gpr_now(GPR_CLOCK_REALTIME),
gpr_time_from_millis(kClientChannelBackupPollIntervalMs * 5 *
poller_slowdown_factor *
grpc_test_slowdown_factor(),
GPR_TIMESPAN)));
SendRpc(stub_.get(), 1, false);
}
TEST_P(End2endTest, RequestStreamOneRequest) {
ResetStub();
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
EchoRequest request;
EchoResponse response;
ClientContext context;
auto stream = stub_->RequestStream(&context, &response);
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
request.set_message("hello");
EXPECT_TRUE(stream->Write(request));
stream->WritesDone();
Status s = stream->Finish();
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
EXPECT_EQ(response.message(), request.message());
EXPECT_TRUE(s.ok());
EXPECT_TRUE(context.debug_error_string().empty());
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
}
TEST_P(End2endTest, RequestStreamOneRequestWithCoalescingApi) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
8 years ago
context.set_initial_metadata_corked(true);
auto stream = stub_->RequestStream(&context, &response);
request.set_message("hello");
stream->WriteLast(request, WriteOptions());
Status s = stream->Finish();
EXPECT_EQ(response.message(), request.message());
EXPECT_TRUE(s.ok());
}
TEST_P(End2endTest, RequestStreamTwoRequests) {
ResetStub();
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
EchoRequest request;
EchoResponse response;
ClientContext context;
auto stream = stub_->RequestStream(&context, &response);
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
request.set_message("hello");
EXPECT_TRUE(stream->Write(request));
EXPECT_TRUE(stream->Write(request));
stream->WritesDone();
Status s = stream->Finish();
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
EXPECT_EQ(response.message(), "hellohello");
EXPECT_TRUE(s.ok());
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
}
TEST_P(End2endTest, RequestStreamTwoRequestsWithWriteThrough) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
auto stream = stub_->RequestStream(&context, &response);
request.set_message("hello");
EXPECT_TRUE(stream->Write(request, WriteOptions().set_write_through()));
EXPECT_TRUE(stream->Write(request, WriteOptions().set_write_through()));
stream->WritesDone();
Status s = stream->Finish();
EXPECT_EQ(response.message(), "hellohello");
EXPECT_TRUE(s.ok());
}
TEST_P(End2endTest, RequestStreamTwoRequestsWithCoalescingApi) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
8 years ago
context.set_initial_metadata_corked(true);
auto stream = stub_->RequestStream(&context, &response);
request.set_message("hello");
EXPECT_TRUE(stream->Write(request));
stream->WriteLast(request, WriteOptions());
Status s = stream->Finish();
EXPECT_EQ(response.message(), "hellohello");
EXPECT_TRUE(s.ok());
}
TEST_P(End2endTest, ResponseStream) {
ResetStub();
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
EchoRequest request;
EchoResponse response;
ClientContext context;
request.set_message("hello");
auto stream = stub_->ResponseStream(&context, request);
for (int i = 0; i < kServerDefaultResponseStreamsToSend; ++i) {
EXPECT_TRUE(stream->Read(&response));
EXPECT_EQ(response.message(), request.message() + std::to_string(i));
}
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
EXPECT_FALSE(stream->Read(&response));
Status s = stream->Finish();
EXPECT_TRUE(s.ok());
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
}
TEST_P(End2endTest, ResponseStreamWithCoalescingApi) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
request.set_message("hello");
context.AddMetadata(kServerUseCoalescingApi, "1");
auto stream = stub_->ResponseStream(&context, request);
for (int i = 0; i < kServerDefaultResponseStreamsToSend; ++i) {
EXPECT_TRUE(stream->Read(&response));
EXPECT_EQ(response.message(), request.message() + std::to_string(i));
}
EXPECT_FALSE(stream->Read(&response));
Status s = stream->Finish();
EXPECT_TRUE(s.ok());
}
// This was added to prevent regression from issue:
// https://github.com/grpc/grpc/issues/11546
TEST_P(End2endTest, ResponseStreamWithEverythingCoalesced) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
request.set_message("hello");
context.AddMetadata(kServerUseCoalescingApi, "1");
// We will only send one message, forcing everything (init metadata, message,
// trailing) to be coalesced together.
context.AddMetadata(kServerResponseStreamsToSend, "1");
auto stream = stub_->ResponseStream(&context, request);
EXPECT_TRUE(stream->Read(&response));
EXPECT_EQ(response.message(), request.message() + "0");
EXPECT_FALSE(stream->Read(&response));
Status s = stream->Finish();
EXPECT_TRUE(s.ok());
}
TEST_P(End2endTest, BidiStream) {
ResetStub();
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
EchoRequest request;
EchoResponse response;
ClientContext context;
std::string msg("hello");
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
auto stream = stub_->BidiStream(&context);
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
for (int i = 0; i < kServerDefaultResponseStreamsToSend; ++i) {
request.set_message(msg + std::to_string(i));
EXPECT_TRUE(stream->Write(request));
EXPECT_TRUE(stream->Read(&response));
EXPECT_EQ(response.message(), request.message());
}
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
stream->WritesDone();
EXPECT_FALSE(stream->Read(&response));
EXPECT_FALSE(stream->Read(&response));
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
Status s = stream->Finish();
EXPECT_TRUE(s.ok());
}
TEST_P(End2endTest, BidiStreamWithCoalescingApi) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
context.AddMetadata(kServerFinishAfterNReads, "3");
8 years ago
context.set_initial_metadata_corked(true);
std::string msg("hello");
auto stream = stub_->BidiStream(&context);
request.set_message(msg + "0");
EXPECT_TRUE(stream->Write(request));
EXPECT_TRUE(stream->Read(&response));
EXPECT_EQ(response.message(), request.message());
request.set_message(msg + "1");
EXPECT_TRUE(stream->Write(request));
EXPECT_TRUE(stream->Read(&response));
EXPECT_EQ(response.message(), request.message());
request.set_message(msg + "2");
stream->WriteLast(request, WriteOptions());
EXPECT_TRUE(stream->Read(&response));
EXPECT_EQ(response.message(), request.message());
EXPECT_FALSE(stream->Read(&response));
EXPECT_FALSE(stream->Read(&response));
Status s = stream->Finish();
EXPECT_TRUE(s.ok());
}
// This was added to prevent regression from issue:
// https://github.com/grpc/grpc/issues/11546
TEST_P(End2endTest, BidiStreamWithEverythingCoalesced) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
context.AddMetadata(kServerFinishAfterNReads, "1");
context.set_initial_metadata_corked(true);
std::string msg("hello");
auto stream = stub_->BidiStream(&context);
request.set_message(msg + "0");
stream->WriteLast(request, WriteOptions());
EXPECT_TRUE(stream->Read(&response));
EXPECT_EQ(response.message(), request.message());
EXPECT_FALSE(stream->Read(&response));
EXPECT_FALSE(stream->Read(&response));
Status s = stream->Finish();
EXPECT_TRUE(s.ok());
}
// Talk to the two services with the same name but different package names.
// The two stubs are created on the same channel.
TEST_P(End2endTest, DiffPackageServices) {
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("Hello");
ClientContext context;
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(response.message(), request.message());
EXPECT_TRUE(s.ok());
std::unique_ptr<grpc::testing::duplicate::EchoTestService::Stub> dup_pkg_stub(
grpc::testing::duplicate::EchoTestService::NewStub(channel_));
ClientContext context2;
s = dup_pkg_stub->Echo(&context2, request, &response);
EXPECT_EQ("no package", response.message());
EXPECT_TRUE(s.ok());
Updating moe_db.txt with the latest equivalence since the ruby import changed the exported structure. Change on 2014/12/01 by nnoble <nnoble@google.com> ------------- new [] file for grpc testing. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Fix unfinished calls in thread_stress_test. Previously we had an early return if we cancelled a stream part way through a message. Correct this, so that half close and full close signals are propagated up the stack correctly so that higher level state machines can see the termination. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Remove dependency on internal C code. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Turn off the flaky bit from thread_stress_test. Change on 2014/12/02 by ctiller <ctiller@google.com> ------------- Add test cases of empty request/response, request streaming, response streaming, and half duplex streaming. Bring up the GFE/ESF for mannual test: [] build java/com/google/net/[]/testing/integration/hexa:server_components_env []-bin/java/com/google/net/[]/testing/integration/hexa/server_components_env --manual --rpc_port=25000 --use_autobahn Change on 2014/12/02 by chenw <chenw@google.com> ------------- Make echo/server.c and fling/server.c shutdown cleanly on SIGINT, and update the relevant tests to exercise this mechanism. Now "[] coverage" and the memory leak detector are able to see into the server processes. Change on 2014/12/02 by pmarks <pmarks@google.com> ------------- Allow the # of channels to be configurable in this performance test. The threads will use the channels in statically-defined round-robin order (not based on when RPCs complete on any channel). The interesting cases are #channels=1 or #channels=#threads (we previously only had the latter case) Change on 2014/12/02 by vpai <vpai@google.com> ------------- Fixed a typo and reworded a comment. Change on 2014/12/02 by gnezdo <gnezdo@google.com> ------------- Require the grpc_call in this ClientContext to be NULL before allowing set_call to be invoked. Otherwise, it's an indication of a leak somewhere. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Correctly return status other than ok and add a test for it. Change on 2014/12/02 by yangg <yangg@google.com> ------------- Better C++ guards for grpc_security.h Change on 2014/12/02 by nnoble <nnoble@google.com> ------------- Use nullptr instead of NULL for consistency. Change on 2014/12/02 by vpai <vpai@google.com> ------------- Updates the ruby gRPC service class to require the serialization method to be a static method - this brings it inline with the proto3 ruby API - it adds a monkey patch to allow existing proto (beefcake) to continue working. Change on 2014/12/02 by temiola <temiola@google.com> ------------- Adding a buildable unit to the blue print file. Added the buildable unit as its name will be usesd as tap project id. This test will fail right away in tap until tests are actually added. Change on 2014/12/02 by donnadionne <donnadionne@google.com> ------------- Move interop ESF C++ server from Java to grpc directory. Tests passed: [] test javatests/com/google/net/[]/testing/integration/hexa/... [] test net/grpc/testing/interop/esf_server/... Change on 2014/12/02 by chenw <chenw@google.com> ------------- Return a lame channel as opposed to NULL when secure channel creation fails. - Looks like we're going to need something similar server-side. - I changed the prototype of the lame client channel factory to take an explicit void as I think this is better practice in C. Let me know if you disagree and I will revert these changes. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- Putting ALPN support where it belongs. Change on 2014/12/02 by jboeuf <jboeuf@google.com> ------------- GOAWAY send path. Sends a GOAWAY frame when shutting down. This is not read and understood yet. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Adds support for secure channels and servers. - wraps new C apis (credentials, server_credentials) and Server#add_secure_http_port - adds tests to ensure credentials and server credentials can be created - updates client_server_spec to run the client_server wrapper layer end-to-end tests using a secure channel Change on 2014/12/03 by temiola <temiola@google.com> ------------- Fix existing issues regarding out of order events. At the client side, using pluck as the client_metadata_read can happen anytime after invoke. At the server side, allow halfclose_ok and rpc_end to come in reverse order. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Don't track coverage of tests. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Change UnaryCall to conform standard test requirement of LargeUnaryCall. Change on 2014/12/03 by yangg <yangg@google.com> ------------- updating alpn version to h2-15 ensure all interop are on the same version and working. Java and go are not ready for h2-16 yet. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Add config to bring echo server in []. This is used to test production GFE as its bckend. Change on 2014/12/03 by chenw <chenw@google.com> ------------- In preparation for fixing shutdown race issues, change em to take ownership of the file descriptor. Add an API to grpc_tcp to take an already created grpc_em_fd object, and change tcp_client to use that API. This is needed because otherwise an em user's close() of the file descriptor may race with libevent internals. That's not an issue yet because destroy() frees the events inline, but that can't be done safely if there is a concurrent poller. Change on 2014/12/03 by klempner <klempner@google.com> ------------- Fixing TAP opensource build We don't want to compile and run C++ tests in the C target. Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Move and separate interop tests by languages. Small fixes to the test runner. Improving logging. Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Fixing the opensource build: -) The C/C++ split wasn't done up to the 'dep' target level -) The alpn.c file was missing from build.json Change on 2014/12/03 by nnoble <nnoble@google.com> ------------- Adding blue print files after projects exist Change on 2014/12/03 by donnadionne <donnadionne@google.com> ------------- Refactor StreamContext using the new completion_queue_pluck API. The dedicated the poller thread has been removed. This CL keeps the current behavior to make it short. There is one following to make it usable for both client and server. The tags for pluck is based on the address of this StreamContext object for potential debug use. The Read/Write and Wait cannot be called concurrently now and this might need to be fixed. Change on 2014/12/03 by yangg <yangg@google.com> ------------- Binary encoding utilities. Support base64 encoding, HPACK static huffman encoding, and doing both at once. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Enforce Makefile regeneration in presubmits. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Make CloseSend() send a real zero-length control message to indicate EOS. Change on 2014/12/03 by zhaoq <zhaoq@google.com> ------------- Prefer to create dualstack sockets for TCP clients and servers, with automatic fallback for environments where IPV6_V6ONLY can't be turned off. Change on 2014/12/03 by pmarks <pmarks@google.com> ------------- Add opensource path to build targets. Ensure that MOE is going to run. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- Add PingPong test case. Delete FullDuplex test case. The latter is not specified for client in https://docs.google.com/document/d/1dwrPpIu5EqiKVsquZfoOqTj7vP8fa1i49gornJo50Qw/edit# Change on 2014/12/03 by chenw <chenw@google.com> ------------- Make generate_projects.sh check out the generated targets. Change on 2014/12/03 by ctiller <ctiller@google.com> ------------- rspec cleanup - stops declaring specs within the GRPC module - splits Bidi streaming specs into a separate test suite adding tests in the GRPC module was a mistake, it pollutes the module and can affect other tests that run later by the test runner the bidi tests are currently flaky, having them run in their own test suite allows having two separate continuous builds (once ruby gRPC is on GitHub), one that includes bidi where we tolerate flakiness, and another that does not, where there should be no flakiness at all Change on 2014/12/03 by temiola <temiola@google.com> ------------- Adding support for composite and IAM credentials. - For now, we don't do any checks on credentials compatibility in the composite credentials. Maybe we'll add that later. - Refactored the end to end security tests so that we always use the public API (except for the fake security context which is not exposed). Change on 2014/12/03 by jboeuf <jboeuf@google.com> ------------- Make GPR library buildable in Visual Studio 2013. Change on 2014/12/04 by jtattermusch <jtattermusch@google.com> ------------- Adds codegen for ruby This is being added now that ruby's proto and grpc apis are defined and stable Change on 2014/12/04 by temiola <temiola@google.com> ------------- Prevent NewStream() from sending negative or 0 timeout. Change on 2014/12/04 by zhaoq <zhaoq@google.com> ------------- Add a grpc_sockaddr_to_string() function, and use it when logging bind failures. Also improve const-correctness in some earlier code. I'm not certain whether inet_ntop() will need any platform-specific implementations, but for now the compiler offers no complaints. Demo: $ []-bin/net/grpc/c/echo_server 1.2.3.4:80 ... tcp_server.c:139] bind addr=[::ffff:1.2.3.4]:80: Permission denied Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Refactoring - moves c wrapped classes to a submodule Google::RPC::Core - this allows for an explicit rule when reading through gRPC ruby code for telling when an object is pure ruby or wrapped C Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixes the bidi_call [] Change on 2014/12/04 by temiola <temiola@google.com> ------------- Fixing dev build when activating surface traces. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Updates the tests to reflect that fact that some Credentials compose works. Change on 2014/12/04 by temiola <temiola@google.com> ------------- Making the generate_project_test actually do something. Change on 2014/12/04 by nnoble <nnoble@google.com> ------------- Rename "esf_server" to "[]4_server". Delete "test_sever" from Java directory. Change on 2014/12/04 by chenw <chenw@google.com> ------------- Added PHP client interop tests. Tested large_unary against the C++ server. Change on 2014/12/04 by mlumish <mlumish@google.com> ------------- Refactor grpc_create_dualstack_socket() by pulling the setsockopt into its own function. This separates the magic test flag from the real fallback logic. Change on 2014/12/04 by pmarks <pmarks@google.com> ------------- Fixes the type of the constant used for test cert hostname Change on 2014/12/04 by temiola <temiola@google.com> ------------- Disabling these tests as they're causing flakiness. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Change intptr --> uintptr. Handles the case where a void* turns into a negative number, which then gets hashed into a negative bucket and segfaults. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Add a test fixture to force parsers to handle one byte at a time. This should expand coverage and hopefully prevent errors at some point (it seems to pass out of the box for now though). Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- The code generator isn't +x. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Updates math_client and math_server to allow construction using crednetials By: - Extending rpc_server constructor so that it takes a credentials keyword param - Extending client_stub constructor so that it takes a credentials keyword param Change on 2014/12/04 by temiola <temiola@google.com> ------------- Format output a little more nicely. Print each line of output separately - previously logging.info was truncating this at some maximum length, and logs were getting lost. Change on 2014/12/04 by ctiller <ctiller@google.com> ------------- Up timeout for this test. Under TSAN, if we process one byte at a time, this timeout can be reached - and I think this is the cause of the following flake: [] Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Adding more error logging for ssl. Change on 2014/12/05 by jboeuf <jboeuf@google.com> ------------- Read path for goaway. Still need to add hooks to deprecate a channel on the client side when goaway is received. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Separate accept() into server_accept() and server_end_of_initial_metadata(). This allows servers to initiate reads before finishing writing metadata. Change on 2014/12/05 by ctiller <ctiller@google.com> ------------- Fix for breakage 11512317 - adding missing test files. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- grpc c++ server side streaming support. This is based on [] There is a lot of room to clean up the internal implementation which may require refactoring of CompletionQueue. The current cl serves as a working implementation with the missing interfaces. The sample generated files are included and will be removed before submitting. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Changed to the latest timeout format again (search "grpc-timeout" in [] for the spec). Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Fixing opensource build. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Making absolutely sure we can do the moe export by adding a sh_test for it. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Change :scheme psuedo-header from "grpc" to "http" or "https". Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Add server credential wrapping for c++ server. It only wraps ssl and []2 for now. The ServerCredentials class and the factory class are in a similar fashion as client side wrapping. The difference is the factory method returns shared_ptr instead of unique_ptr as the server builder needs to keep a reference to it for actually creating the server later. The integration will happen in a following cl. Change on 2014/12/05 by yangg <yangg@google.com> ------------- Fixed bugs in new_grpc_docker_builder.sh Change on 2014/12/05 by mlumish <mlumish@google.com> ------------- In secure endpoint, hold a refcount for the life of a write callback if the write does not complete immediately. Change on 2014/12/05 by klempner <klempner@google.com> ------------- Add migration support to MOE and have TAP verify it doesn't break. Migration support allows mirroring commits from [] into the git repo, instead of just a dump of the current source. Change on 2014/12/05 by ejona <ejona@google.com> ------------- Change initial window size to 65535 according http2 draft 15. Change on 2014/12/05 by zhaoq <zhaoq@google.com> ------------- Re-enable the flaky cases in dualstack_socket_test, with additional logging to help track down the problem if it surfaces again. This also seems like a good opportunity to make grpc_socket_utils a separate library, as it's not really specific to TCP. Example output: logspam: [], 26570) resolved 2 addrs in 37ms: logspam: [0] [::1]:26570 logspam: [1] 127.0.0.1:26570 Change on 2014/12/05 by pmarks <pmarks@google.com> ------------- Opensource build fixes. -) A function that has a return type should actually return something. -) Don't pass unsigned chars to strlen and strncmp. Change on 2014/12/05 by nnoble <nnoble@google.com> ------------- Created by MOE: http://code.google.com/p/moe-java MOE_MIGRATED_REVID=81458281
10 years ago
}
template <class ServiceType>
void CancelRpc(ClientContext* context, int delay_us, ServiceType* service) {
10 years ago
gpr_sleep_until(gpr_time_add(gpr_now(GPR_CLOCK_REALTIME),
10 years ago
gpr_time_from_micros(delay_us, GPR_TIMESPAN)));
while (!service->signal_client()) {
}
context->TryCancel();
}
TEST_P(End2endTest, CancelRpcBeforeStart) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
request.set_message("hello");
context.TryCancel();
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ("", response.message());
EXPECT_EQ(grpc::StatusCode::CANCELLED, s.error_code());
if (GetParam().use_interceptors()) {
EXPECT_EQ(20, PhonyInterceptor::GetNumTimesCancel());
}
}
TEST_P(End2endTest, CancelRpcAfterStart) {
[promises] Run C++ end to end tests with server promises (#32537) Expand server promises to run with C++ end2end tests. Across connected_channel/call/batch_builder/pipe/transport: - fix a bug where read errors weren't propagated from transport to call so that we can populate failed_before_recv_message for the c++ bindings - ensure those errors are not, however, used to populate the returned call status Add a new latch call arg to lazily propagate the bound CQ for a server call (and client call, but here it's used degenerately - it's always populated). This allows server calls to be properly bound to pollsets.(1)/(2) In call.cc: - move some profiling code from FilterStackCall to Call, and then use it in PromiseBasedCall (this should be cleaned up with tracing work) - implement GetServerAuthority In server.cc: - use an RAII pattern on `MatchResult` to avoid a bug whereby a tag could be dropped if we cancel a request after it's been matched but before it's published - fix deadline export to ServerContext In resource_quota_server.cc: - fix some long standing flakes (that were finally obvious with the new test code) - it's legal here to have client calls not arrive at the server due to resource starvation, work through that (includes adding expectations during a `Step` call, which required some small tweaks to cq_verifier) In the C++ end2end_test.cc: - strengthen a flaky test so it passes consistently (it's likely we'll revisit this with the fuzzing efforts to strengthen it into an actually robust test) (1) It's time to remove this concept (2) Surprisingly the only test that *reliably* demonstrates this not being done is time_change_test --------- Co-authored-by: ctiller <ctiller@users.noreply.github.com>
2 years ago
for (int i = 0; i < 10; i++) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
request.set_message("hello");
request.mutable_param()->set_server_notify_client_when_started(true);
request.mutable_param()->set_skip_cancelled_check(true);
Status s;
std::thread echo_thread([this, &s, &context, &request, &response] {
s = stub_->Echo(&context, request, &response);
});
if (!GetParam().callback_server()) {
EXPECT_EQ(service_.ClientWaitUntilNRpcsStarted(1), 1);
[promises] Run C++ end to end tests with server promises (#32537) Expand server promises to run with C++ end2end tests. Across connected_channel/call/batch_builder/pipe/transport: - fix a bug where read errors weren't propagated from transport to call so that we can populate failed_before_recv_message for the c++ bindings - ensure those errors are not, however, used to populate the returned call status Add a new latch call arg to lazily propagate the bound CQ for a server call (and client call, but here it's used degenerately - it's always populated). This allows server calls to be properly bound to pollsets.(1)/(2) In call.cc: - move some profiling code from FilterStackCall to Call, and then use it in PromiseBasedCall (this should be cleaned up with tracing work) - implement GetServerAuthority In server.cc: - use an RAII pattern on `MatchResult` to avoid a bug whereby a tag could be dropped if we cancel a request after it's been matched but before it's published - fix deadline export to ServerContext In resource_quota_server.cc: - fix some long standing flakes (that were finally obvious with the new test code) - it's legal here to have client calls not arrive at the server due to resource starvation, work through that (includes adding expectations during a `Step` call, which required some small tweaks to cq_verifier) In the C++ end2end_test.cc: - strengthen a flaky test so it passes consistently (it's likely we'll revisit this with the fuzzing efforts to strengthen it into an actually robust test) (1) It's time to remove this concept (2) Surprisingly the only test that *reliably* demonstrates this not being done is time_change_test --------- Co-authored-by: ctiller <ctiller@users.noreply.github.com>
2 years ago
} else {
EXPECT_EQ(callback_service_.ClientWaitUntilNRpcsStarted(1), 1);
[promises] Run C++ end to end tests with server promises (#32537) Expand server promises to run with C++ end2end tests. Across connected_channel/call/batch_builder/pipe/transport: - fix a bug where read errors weren't propagated from transport to call so that we can populate failed_before_recv_message for the c++ bindings - ensure those errors are not, however, used to populate the returned call status Add a new latch call arg to lazily propagate the bound CQ for a server call (and client call, but here it's used degenerately - it's always populated). This allows server calls to be properly bound to pollsets.(1)/(2) In call.cc: - move some profiling code from FilterStackCall to Call, and then use it in PromiseBasedCall (this should be cleaned up with tracing work) - implement GetServerAuthority In server.cc: - use an RAII pattern on `MatchResult` to avoid a bug whereby a tag could be dropped if we cancel a request after it's been matched but before it's published - fix deadline export to ServerContext In resource_quota_server.cc: - fix some long standing flakes (that were finally obvious with the new test code) - it's legal here to have client calls not arrive at the server due to resource starvation, work through that (includes adding expectations during a `Step` call, which required some small tweaks to cq_verifier) In the C++ end2end_test.cc: - strengthen a flaky test so it passes consistently (it's likely we'll revisit this with the fuzzing efforts to strengthen it into an actually robust test) (1) It's time to remove this concept (2) Surprisingly the only test that *reliably* demonstrates this not being done is time_change_test --------- Co-authored-by: ctiller <ctiller@users.noreply.github.com>
2 years ago
}
[promises] Run C++ end to end tests with server promises (#32537) Expand server promises to run with C++ end2end tests. Across connected_channel/call/batch_builder/pipe/transport: - fix a bug where read errors weren't propagated from transport to call so that we can populate failed_before_recv_message for the c++ bindings - ensure those errors are not, however, used to populate the returned call status Add a new latch call arg to lazily propagate the bound CQ for a server call (and client call, but here it's used degenerately - it's always populated). This allows server calls to be properly bound to pollsets.(1)/(2) In call.cc: - move some profiling code from FilterStackCall to Call, and then use it in PromiseBasedCall (this should be cleaned up with tracing work) - implement GetServerAuthority In server.cc: - use an RAII pattern on `MatchResult` to avoid a bug whereby a tag could be dropped if we cancel a request after it's been matched but before it's published - fix deadline export to ServerContext In resource_quota_server.cc: - fix some long standing flakes (that were finally obvious with the new test code) - it's legal here to have client calls not arrive at the server due to resource starvation, work through that (includes adding expectations during a `Step` call, which required some small tweaks to cq_verifier) In the C++ end2end_test.cc: - strengthen a flaky test so it passes consistently (it's likely we'll revisit this with the fuzzing efforts to strengthen it into an actually robust test) (1) It's time to remove this concept (2) Surprisingly the only test that *reliably* demonstrates this not being done is time_change_test --------- Co-authored-by: ctiller <ctiller@users.noreply.github.com>
2 years ago
context.TryCancel();
[promises] Run C++ end to end tests with server promises (#32537) Expand server promises to run with C++ end2end tests. Across connected_channel/call/batch_builder/pipe/transport: - fix a bug where read errors weren't propagated from transport to call so that we can populate failed_before_recv_message for the c++ bindings - ensure those errors are not, however, used to populate the returned call status Add a new latch call arg to lazily propagate the bound CQ for a server call (and client call, but here it's used degenerately - it's always populated). This allows server calls to be properly bound to pollsets.(1)/(2) In call.cc: - move some profiling code from FilterStackCall to Call, and then use it in PromiseBasedCall (this should be cleaned up with tracing work) - implement GetServerAuthority In server.cc: - use an RAII pattern on `MatchResult` to avoid a bug whereby a tag could be dropped if we cancel a request after it's been matched but before it's published - fix deadline export to ServerContext In resource_quota_server.cc: - fix some long standing flakes (that were finally obvious with the new test code) - it's legal here to have client calls not arrive at the server due to resource starvation, work through that (includes adding expectations during a `Step` call, which required some small tweaks to cq_verifier) In the C++ end2end_test.cc: - strengthen a flaky test so it passes consistently (it's likely we'll revisit this with the fuzzing efforts to strengthen it into an actually robust test) (1) It's time to remove this concept (2) Surprisingly the only test that *reliably* demonstrates this not being done is time_change_test --------- Co-authored-by: ctiller <ctiller@users.noreply.github.com>
2 years ago
if (!GetParam().callback_server()) {
service_.SignalServerToContinue();
} else {
callback_service_.SignalServerToContinue();
}
[promises] Run C++ end to end tests with server promises (#32537) Expand server promises to run with C++ end2end tests. Across connected_channel/call/batch_builder/pipe/transport: - fix a bug where read errors weren't propagated from transport to call so that we can populate failed_before_recv_message for the c++ bindings - ensure those errors are not, however, used to populate the returned call status Add a new latch call arg to lazily propagate the bound CQ for a server call (and client call, but here it's used degenerately - it's always populated). This allows server calls to be properly bound to pollsets.(1)/(2) In call.cc: - move some profiling code from FilterStackCall to Call, and then use it in PromiseBasedCall (this should be cleaned up with tracing work) - implement GetServerAuthority In server.cc: - use an RAII pattern on `MatchResult` to avoid a bug whereby a tag could be dropped if we cancel a request after it's been matched but before it's published - fix deadline export to ServerContext In resource_quota_server.cc: - fix some long standing flakes (that were finally obvious with the new test code) - it's legal here to have client calls not arrive at the server due to resource starvation, work through that (includes adding expectations during a `Step` call, which required some small tweaks to cq_verifier) In the C++ end2end_test.cc: - strengthen a flaky test so it passes consistently (it's likely we'll revisit this with the fuzzing efforts to strengthen it into an actually robust test) (1) It's time to remove this concept (2) Surprisingly the only test that *reliably* demonstrates this not being done is time_change_test --------- Co-authored-by: ctiller <ctiller@users.noreply.github.com>
2 years ago
echo_thread.join();
// TODO(ctiller): improve test to not be flaky
//
// TryCancel is best effort, and it can happen that the cancellation is not
// acted upon before the server wakes up, sends a response, and the client
// reads that.
// For this reason, we try a few times here to see the cancellation result.
if (s.ok()) continue;
EXPECT_EQ("", response.message());
EXPECT_EQ(grpc::StatusCode::CANCELLED, s.error_code());
if (GetParam().use_interceptors()) {
EXPECT_EQ(20, PhonyInterceptor::GetNumTimesCancel());
}
return;
}
[promises] Run C++ end to end tests with server promises (#32537) Expand server promises to run with C++ end2end tests. Across connected_channel/call/batch_builder/pipe/transport: - fix a bug where read errors weren't propagated from transport to call so that we can populate failed_before_recv_message for the c++ bindings - ensure those errors are not, however, used to populate the returned call status Add a new latch call arg to lazily propagate the bound CQ for a server call (and client call, but here it's used degenerately - it's always populated). This allows server calls to be properly bound to pollsets.(1)/(2) In call.cc: - move some profiling code from FilterStackCall to Call, and then use it in PromiseBasedCall (this should be cleaned up with tracing work) - implement GetServerAuthority In server.cc: - use an RAII pattern on `MatchResult` to avoid a bug whereby a tag could be dropped if we cancel a request after it's been matched but before it's published - fix deadline export to ServerContext In resource_quota_server.cc: - fix some long standing flakes (that were finally obvious with the new test code) - it's legal here to have client calls not arrive at the server due to resource starvation, work through that (includes adding expectations during a `Step` call, which required some small tweaks to cq_verifier) In the C++ end2end_test.cc: - strengthen a flaky test so it passes consistently (it's likely we'll revisit this with the fuzzing efforts to strengthen it into an actually robust test) (1) It's time to remove this concept (2) Surprisingly the only test that *reliably* demonstrates this not being done is time_change_test --------- Co-authored-by: ctiller <ctiller@users.noreply.github.com>
2 years ago
GTEST_FAIL() << "Failed to get cancellation";
}
// Client cancels request stream after sending two messages
TEST_P(End2endTest, ClientCancelsRequestStream) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
request.set_message("hello");
auto stream = stub_->RequestStream(&context, &response);
EXPECT_TRUE(stream->Write(request));
EXPECT_TRUE(stream->Write(request));
10 years ago
context.TryCancel();
Status s = stream->Finish();
EXPECT_EQ(grpc::StatusCode::CANCELLED, s.error_code());
10 years ago
EXPECT_EQ(response.message(), "");
if (GetParam().use_interceptors()) {
EXPECT_EQ(20, PhonyInterceptor::GetNumTimesCancel());
}
}
// Client cancels server stream after sending some messages
TEST_P(End2endTest, ClientCancelsResponseStream) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
request.set_message("hello");
auto stream = stub_->ResponseStream(&context, request);
EXPECT_TRUE(stream->Read(&response));
EXPECT_EQ(response.message(), request.message() + "0");
EXPECT_TRUE(stream->Read(&response));
EXPECT_EQ(response.message(), request.message() + "1");
context.TryCancel();
// The cancellation races with responses, so there might be zero or
// one responses pending, read till failure
if (stream->Read(&response)) {
EXPECT_EQ(response.message(), request.message() + "2");
// Since we have cancelled, we expect the next attempt to read to fail
EXPECT_FALSE(stream->Read(&response));
}
Status s = stream->Finish();
// The final status could be either of CANCELLED or OK depending on
// who won the race.
EXPECT_GE(grpc::StatusCode::CANCELLED, s.error_code());
if (GetParam().use_interceptors()) {
EXPECT_EQ(20, PhonyInterceptor::GetNumTimesCancel());
}
}
// Client cancels bidi stream after sending some messages
TEST_P(End2endTest, ClientCancelsBidi) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
std::string msg("hello");
// Send server_try_cancel value in the client metadata
context.AddMetadata(kClientTryCancelRequest, std::to_string(1));
auto stream = stub_->BidiStream(&context);
request.set_message(msg + "0");
EXPECT_TRUE(stream->Write(request));
EXPECT_TRUE(stream->Read(&response));
EXPECT_EQ(response.message(), request.message());
request.set_message(msg + "1");
EXPECT_TRUE(stream->Write(request));
context.TryCancel();
// The cancellation races with responses, so there might be zero or
// one responses pending, read till failure
if (stream->Read(&response)) {
EXPECT_EQ(response.message(), request.message());
// Since we have cancelled, we expect the next attempt to read to fail
EXPECT_FALSE(stream->Read(&response));
}
Status s = stream->Finish();
EXPECT_EQ(grpc::StatusCode::CANCELLED, s.error_code());
if (GetParam().use_interceptors()) {
EXPECT_EQ(20, PhonyInterceptor::GetNumTimesCancel());
}
}
TEST_P(End2endTest, RpcMaxMessageSize) {
ResetStub();
EchoRequest request;
EchoResponse response;
10 years ago
request.set_message(string(kMaxMessageSize_ * 2, 'a'));
request.mutable_param()->set_server_die(true);
ClientContext context;
Status s = stub_->Echo(&context, request, &response);
EXPECT_FALSE(s.ok());
}
void ReaderThreadFunc(ClientReaderWriter<EchoRequest, EchoResponse>* stream,
gpr_event* ev) {
EchoResponse resp;
gpr_event_set(ev, reinterpret_cast<void*>(1));
while (stream->Read(&resp)) {
LOG(INFO) << "Read message";
}
}
// Run a Read and a WritesDone simultaneously.
TEST_P(End2endTest, SimultaneousReadWritesDone) {
ResetStub();
ClientContext context;
gpr_event ev;
gpr_event_init(&ev);
auto stream = stub_->BidiStream(&context);
std::thread reader_thread(ReaderThreadFunc, stream.get(), &ev);
gpr_event_wait(&ev, gpr_inf_future(GPR_CLOCK_REALTIME));
stream->WritesDone();
reader_thread.join();
Status s = stream->Finish();
EXPECT_TRUE(s.ok());
}
TEST_P(End2endTest, ChannelState) {
if (GetParam().inproc()) {
return;
}
ResetStub();
// Start IDLE
EXPECT_EQ(GRPC_CHANNEL_IDLE, channel_->GetState(false));
// Did not ask to connect, no state change.
CompletionQueue cq;
std::chrono::system_clock::time_point deadline =
std::chrono::system_clock::now() + std::chrono::milliseconds(10);
channel_->NotifyOnStateChange(GRPC_CHANNEL_IDLE, deadline, &cq, nullptr);
void* tag;
bool ok = true;
cq.Next(&tag, &ok);
EXPECT_FALSE(ok);
EXPECT_EQ(GRPC_CHANNEL_IDLE, channel_->GetState(true));
EXPECT_TRUE(channel_->WaitForStateChange(GRPC_CHANNEL_IDLE,
gpr_inf_future(GPR_CLOCK_REALTIME)));
auto state = channel_->GetState(false);
EXPECT_TRUE(state == GRPC_CHANNEL_CONNECTING || state == GRPC_CHANNEL_READY);
}
// Takes 10s.
TEST_P(End2endTest, ChannelStateTimeout) {
if ((GetParam().credentials_type() != kInsecureCredentialsType) ||
GetParam().inproc()) {
return;
}
int port = grpc_pick_unused_port_or_die();
std::ostringstream server_address;
server_address << "localhost:" << port;
// Channel to non-existing server
auto channel =
grpc::CreateChannel(server_address.str(), InsecureChannelCredentials());
// Start IDLE
EXPECT_EQ(GRPC_CHANNEL_IDLE, channel->GetState(true));
auto state = GRPC_CHANNEL_IDLE;
for (int i = 0; i < 10; i++) {
channel->WaitForStateChange(
state, std::chrono::system_clock::now() + std::chrono::seconds(1));
state = channel->GetState(false);
}
}
TEST_P(End2endTest, ChannelStateOnLameChannel) {
if ((GetParam().credentials_type() != kInsecureCredentialsType) ||
GetParam().inproc()) {
return;
}
// Channel using invalid target URI. This creates a lame channel.
auto channel = grpc::CreateChannel("dns:///", InsecureChannelCredentials());
// Channel should immediately report TRANSIENT_FAILURE.
EXPECT_EQ(GRPC_CHANNEL_TRANSIENT_FAILURE, channel->GetState(true));
// And state will never change.
auto state = GRPC_CHANNEL_TRANSIENT_FAILURE;
for (int i = 0; i < 10; ++i) {
channel->WaitForStateChange(
state, std::chrono::system_clock::now() + std::chrono::seconds(1));
state = channel->GetState(false);
}
}
// Talking to a non-existing service.
TEST_P(End2endTest, NonExistingService) {
ResetChannel();
std::unique_ptr<grpc::testing::UnimplementedEchoService::Stub> stub;
stub = grpc::testing::UnimplementedEchoService::NewStub(channel_);
EchoRequest request;
EchoResponse response;
request.set_message("Hello");
ClientContext context;
Status s = stub->Unimplemented(&context, request, &response);
EXPECT_EQ(StatusCode::UNIMPLEMENTED, s.error_code());
EXPECT_EQ("", s.error_message());
}
// Ask the server to send back a serialized proto in trailer.
// This is an example of setting error details.
TEST_P(End2endTest, BinaryTrailerTest) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
request.mutable_param()->set_echo_metadata(true);
DebugInfo* info = request.mutable_param()->mutable_debug_info();
info->add_stack_entries("stack_entry_1");
info->add_stack_entries("stack_entry_2");
info->add_stack_entries("stack_entry_3");
info->set_detail("detailed debug info");
std::string expected_string = info->SerializeAsString();
request.set_message("Hello");
Status s = stub_->Echo(&context, request, &response);
EXPECT_FALSE(s.ok());
auto trailers = context.GetServerTrailingMetadata();
EXPECT_EQ(1u, trailers.count(kDebugInfoTrailerKey));
auto iter = trailers.find(kDebugInfoTrailerKey);
EXPECT_EQ(expected_string, iter->second);
// Parse the returned trailer into a DebugInfo proto.
DebugInfo returned_info;
EXPECT_TRUE(returned_info.ParseFromString(ToString(iter->second)));
}
TEST_P(End2endTest, ExpectErrorTest) {
ResetStub();
std::vector<ErrorStatus> expected_status;
7 years ago
expected_status.emplace_back();
expected_status.back().set_code(13); // INTERNAL
// No Error message or details
expected_status.emplace_back();
expected_status.back().set_code(13); // INTERNAL
expected_status.back().set_error_message("text error message");
expected_status.back().set_binary_error_details("text error details");
7 years ago
expected_status.emplace_back();
expected_status.back().set_code(13); // INTERNAL
expected_status.back().set_error_message("text error message");
expected_status.back().set_binary_error_details(
"\x0\x1\x2\x3\x4\x5\x6\x8\x9\xA\xB");
for (auto iter = expected_status.begin(); iter != expected_status.end();
++iter) {
EchoRequest request;
EchoResponse response;
ClientContext context;
request.set_message("Hello");
auto* error = request.mutable_param()->mutable_expected_error();
error->set_code(iter->code());
error->set_error_message(iter->error_message());
error->set_binary_error_details(iter->binary_error_details());
Status s = stub_->Echo(&context, request, &response);
EXPECT_FALSE(s.ok());
EXPECT_EQ(iter->code(), s.error_code());
EXPECT_EQ(iter->error_message(), s.error_message());
EXPECT_EQ(iter->binary_error_details(), s.error_details());
EXPECT_TRUE(absl::StrContains(context.debug_error_string(), "created"));
#ifndef NDEBUG
// grpc_core::StatusIntProperty::kFileLine is for debug only
EXPECT_TRUE(absl::StrContains(context.debug_error_string(), "file"));
EXPECT_TRUE(absl::StrContains(context.debug_error_string(), "line"));
#endif
EXPECT_TRUE(absl::StrContains(context.debug_error_string(), "status"));
EXPECT_TRUE(absl::StrContains(context.debug_error_string(), "13"));
}
}
//////////////////////////////////////////////////////////////////////////
// Test with and without a proxy.
class ProxyEnd2endTest : public End2endTest {
protected:
};
TEST_P(ProxyEnd2endTest, SimpleRpc) {
ResetStub();
SendRpc(stub_.get(), 1, false);
}
TEST_P(ProxyEnd2endTest, SimpleRpcWithEmptyMessages) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
Status s = stub_->Echo(&context, request, &response);
EXPECT_TRUE(s.ok());
}
TEST_P(ProxyEnd2endTest, MultipleRpcs) {
ResetStub();
std::vector<std::thread> threads;
threads.reserve(10);
for (int i = 0; i < 10; ++i) {
threads.emplace_back(SendRpc, stub_.get(), 10, false);
}
for (int i = 0; i < 10; ++i) {
threads[i].join();
}
}
// Set a 10us deadline and make sure proper error is returned.
TEST_P(ProxyEnd2endTest, RpcDeadlineExpires) {
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("Hello");
request.mutable_param()->set_skip_cancelled_check(true);
// Let server sleep for 4 secs first to guarantee expiry.
// 4 secs might seem a bit extreme but the timer manager would have been just
// initialized (when ResetStub() was called) and there are some warmup costs
// i.e the timer thread many not have even started. There might also be other
// delays in the timer manager thread (in acquiring locks, timer data
// structure manipulations, starting backup timer threads) that add to the
// delays. 4 secs might be still not enough in some cases but this
// significantly reduces the test flakes
request.mutable_param()->set_server_sleep_us(4 * 1000 * 1000);
ClientContext context;
std::chrono::system_clock::time_point deadline =
std::chrono::system_clock::now() + std::chrono::milliseconds(1);
context.set_deadline(deadline);
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(StatusCode::DEADLINE_EXCEEDED, s.error_code());
}
// Set a long but finite deadline.
TEST_P(ProxyEnd2endTest, RpcLongDeadline) {
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("Hello");
ClientContext context;
std::chrono::system_clock::time_point deadline =
std::chrono::system_clock::now() + std::chrono::hours(1);
context.set_deadline(deadline);
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(response.message(), request.message());
EXPECT_TRUE(s.ok());
}
// Ask server to echo back the deadline it sees.
TEST_P(ProxyEnd2endTest, EchoDeadline) {
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("Hello");
request.mutable_param()->set_echo_deadline(true);
ClientContext context;
std::chrono::system_clock::time_point deadline =
std::chrono::system_clock::now() + std::chrono::seconds(100);
context.set_deadline(deadline);
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(response.message(), request.message());
EXPECT_TRUE(s.ok());
gpr_timespec sent_deadline;
Timepoint2Timespec(deadline, &sent_deadline);
// We want to allow some reasonable error given:
// - request_deadline() only has 1sec resolution so the best we can do is +-1
// - if sent_deadline.tv_nsec is very close to the next second's boundary we
// can end up being off by 2 in one direction.
EXPECT_LE(response.param().request_deadline() - sent_deadline.tv_sec, 2);
EXPECT_GE(response.param().request_deadline() - sent_deadline.tv_sec, -1);
}
// Ask server to echo back the deadline it sees. The rpc has no deadline.
TEST_P(ProxyEnd2endTest, EchoDeadlineForNoDeadlineRpc) {
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("Hello");
request.mutable_param()->set_echo_deadline(true);
ClientContext context;
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(response.message(), request.message());
EXPECT_TRUE(s.ok());
EXPECT_EQ(response.param().request_deadline(),
gpr_inf_future(GPR_CLOCK_REALTIME).tv_sec);
}
TEST_P(ProxyEnd2endTest, UnimplementedRpc) {
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("Hello");
ClientContext context;
Status s = stub_->Unimplemented(&context, request, &response);
EXPECT_FALSE(s.ok());
EXPECT_EQ(s.error_code(), grpc::StatusCode::UNIMPLEMENTED);
EXPECT_EQ(s.error_message(), "");
EXPECT_EQ(response.message(), "");
}
// Client cancels rpc after 10ms
TEST_P(ProxyEnd2endTest, ClientCancelsRpc) {
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("Hello");
const int kCancelDelayUs = 10 * 1000;
request.mutable_param()->set_client_cancel_after_us(kCancelDelayUs);
ClientContext context;
std::thread cancel_thread;
if (!GetParam().callback_server()) {
cancel_thread = std::thread(
[&context, this](int delay) { CancelRpc(&context, delay, &service_); },
kCancelDelayUs);
// Note: the unusual pattern above (and below) is caused by a conflict
// between two sets of compiler expectations. clang allows const to be
// captured without mention, so there is no need to capture kCancelDelayUs
// (and indeed clang-tidy complains if you do so). OTOH, a Windows compiler
// in our tests requires an explicit capture even for const. We square this
// circle by passing the const value in as an argument to the lambda.
} else {
cancel_thread = std::thread(
[&context, this](int delay) {
CancelRpc(&context, delay, &callback_service_);
},
kCancelDelayUs);
}
Status s = stub_->Echo(&context, request, &response);
cancel_thread.join();
EXPECT_EQ(StatusCode::CANCELLED, s.error_code());
EXPECT_EQ(s.error_message(), "CANCELLED");
}
// Server cancels rpc after 1ms
TEST_P(ProxyEnd2endTest, ServerCancelsRpc) {
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("Hello");
request.mutable_param()->set_server_cancel_after_us(1000);
ClientContext context;
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(StatusCode::CANCELLED, s.error_code());
EXPECT_TRUE(s.error_message().empty());
}
// Make the response larger than the flow control window.
TEST_P(ProxyEnd2endTest, HugeResponse) {
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("huge response");
const size_t kResponseSize = 1024 * (1024 + 10);
request.mutable_param()->set_response_message_length(kResponseSize);
ClientContext context;
std::chrono::system_clock::time_point deadline =
std::chrono::system_clock::now() + std::chrono::seconds(20);
context.set_deadline(deadline);
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(kResponseSize, response.message().size());
EXPECT_TRUE(s.ok());
}
TEST_P(ProxyEnd2endTest, Peer) {
// Peer is not meaningful for inproc
if (GetParam().inproc()) {
return;
}
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("hello");
request.mutable_param()->set_echo_peer(true);
ClientContext context;
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(response.message(), request.message());
EXPECT_TRUE(s.ok());
EXPECT_TRUE(CheckIsLocalhost(response.param().peer()));
EXPECT_TRUE(CheckIsLocalhost(context.peer()));
}
//////////////////////////////////////////////////////////////////////////
class SecureEnd2endTest : public End2endTest {
protected:
SecureEnd2endTest() {
CHECK(!GetParam().use_proxy());
CHECK(GetParam().credentials_type() != kInsecureCredentialsType);
}
};
TEST_P(SecureEnd2endTest, SimpleRpcWithHost) {
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("Hello");
ClientContext context;
context.set_authority("foo.test.youtube.com");
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(response.message(), request.message());
EXPECT_TRUE(response.has_param());
EXPECT_EQ("special", response.param().host());
EXPECT_TRUE(s.ok());
}
bool MetadataContains(
const std::multimap<grpc::string_ref, grpc::string_ref>& metadata,
const std::string& key, const std::string& value) {
int count = 0;
for (std::multimap<grpc::string_ref, grpc::string_ref>::const_iterator iter =
metadata.begin();
iter != metadata.end(); ++iter) {
if (ToString(iter->first) == key && ToString(iter->second) == value) {
count++;
}
}
return count == 1;
}
TEST_P(SecureEnd2endTest, BlockingAuthMetadataPluginAndProcessorSuccess) {
auto* processor = new TestAuthMetadataProcessor(true);
StartServer(std::shared_ptr<AuthMetadataProcessor>(processor));
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
context.set_credentials(processor->GetCompatibleClientCreds());
request.set_message("Hello");
request.mutable_param()->set_echo_metadata(true);
request.mutable_param()->set_expected_client_identity(
TestAuthMetadataProcessor::kGoodGuy);
request.mutable_param()->set_expected_transport_security_type(
GetParam().credentials_type());
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(request.message(), response.message());
EXPECT_TRUE(s.ok());
// Metadata should have been consumed by the processor.
EXPECT_FALSE(MetadataContains(
context.GetServerTrailingMetadata(), GRPC_AUTHORIZATION_METADATA_KEY,
std::string("Bearer ") + TestAuthMetadataProcessor::kGoodGuy));
}
TEST_P(SecureEnd2endTest, BlockingAuthMetadataPluginAndProcessorFailure) {
auto* processor = new TestAuthMetadataProcessor(true);
StartServer(std::shared_ptr<AuthMetadataProcessor>(processor));
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
context.set_credentials(processor->GetIncompatibleClientCreds());
request.set_message("Hello");
Status s = stub_->Echo(&context, request, &response);
EXPECT_FALSE(s.ok());
EXPECT_EQ(s.error_code(), StatusCode::UNAUTHENTICATED);
}
TEST_P(SecureEnd2endTest, SetPerCallCredentials) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
std::shared_ptr<CallCredentials> creds =
GoogleIAMCredentials(kFakeToken, kFakeSelector);
context.set_credentials(creds);
request.set_message("Hello");
request.mutable_param()->set_echo_metadata(true);
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(request.message(), response.message());
EXPECT_TRUE(s.ok());
EXPECT_TRUE(MetadataContains(context.GetServerTrailingMetadata(),
GRPC_IAM_AUTHORIZATION_TOKEN_METADATA_KEY,
kFakeToken));
EXPECT_TRUE(MetadataContains(context.GetServerTrailingMetadata(),
GRPC_IAM_AUTHORITY_SELECTOR_METADATA_KEY,
kFakeSelector));
EXPECT_EQ(context.credentials()->DebugString(),
kExpectedFakeCredsDebugString);
}
class CredentialsInterceptor : public experimental::Interceptor {
public:
explicit CredentialsInterceptor(experimental::ClientRpcInfo* info)
: info_(info) {}
void Intercept(experimental::InterceptorBatchMethods* methods) override {
if (methods->QueryInterceptionHookPoint(
experimental::InterceptionHookPoints::PRE_SEND_INITIAL_METADATA)) {
std::shared_ptr<CallCredentials> creds =
GoogleIAMCredentials(kFakeToken, kFakeSelector);
info_->client_context()->set_credentials(creds);
}
methods->Proceed();
}
private:
experimental::ClientRpcInfo* info_ = nullptr;
};
class CredentialsInterceptorFactory
: public experimental::ClientInterceptorFactoryInterface {
CredentialsInterceptor* CreateClientInterceptor(
experimental::ClientRpcInfo* info) override {
return new CredentialsInterceptor(info);
}
};
TEST_P(SecureEnd2endTest, CallCredentialsInterception) {
if (!GetParam().use_interceptors()) {
return;
}
std::vector<std::unique_ptr<experimental::ClientInterceptorFactoryInterface>>
interceptor_creators;
interceptor_creators.push_back(
std::make_unique<CredentialsInterceptorFactory>());
ResetStub(std::move(interceptor_creators));
EchoRequest request;
EchoResponse response;
ClientContext context;
request.set_message("Hello");
request.mutable_param()->set_echo_metadata(true);
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(request.message(), response.message());
EXPECT_TRUE(s.ok());
EXPECT_TRUE(MetadataContains(context.GetServerTrailingMetadata(),
GRPC_IAM_AUTHORIZATION_TOKEN_METADATA_KEY,
kFakeToken));
EXPECT_TRUE(MetadataContains(context.GetServerTrailingMetadata(),
GRPC_IAM_AUTHORITY_SELECTOR_METADATA_KEY,
kFakeSelector));
EXPECT_EQ(context.credentials()->DebugString(),
kExpectedFakeCredsDebugString);
}
TEST_P(SecureEnd2endTest, CallCredentialsInterceptionWithSetCredentials) {
if (!GetParam().use_interceptors()) {
return;
}
std::vector<std::unique_ptr<experimental::ClientInterceptorFactoryInterface>>
interceptor_creators;
interceptor_creators.push_back(
std::make_unique<CredentialsInterceptorFactory>());
ResetStub(std::move(interceptor_creators));
EchoRequest request;
EchoResponse response;
ClientContext context;
std::shared_ptr<CallCredentials> creds1 =
GoogleIAMCredentials(kWrongToken, kWrongSelector);
context.set_credentials(creds1);
EXPECT_EQ(context.credentials(), creds1);
EXPECT_EQ(context.credentials()->DebugString(),
kExpectedWrongCredsDebugString);
request.set_message("Hello");
request.mutable_param()->set_echo_metadata(true);
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(request.message(), response.message());
EXPECT_TRUE(s.ok());
EXPECT_TRUE(MetadataContains(context.GetServerTrailingMetadata(),
GRPC_IAM_AUTHORIZATION_TOKEN_METADATA_KEY,
kFakeToken));
EXPECT_TRUE(MetadataContains(context.GetServerTrailingMetadata(),
GRPC_IAM_AUTHORITY_SELECTOR_METADATA_KEY,
kFakeSelector));
EXPECT_EQ(context.credentials()->DebugString(),
kExpectedFakeCredsDebugString);
}
TEST_P(SecureEnd2endTest, OverridePerCallCredentials) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
std::shared_ptr<CallCredentials> creds1 =
GoogleIAMCredentials(kFakeToken1, kFakeSelector1);
context.set_credentials(creds1);
EXPECT_EQ(context.credentials(), creds1);
EXPECT_EQ(context.credentials()->DebugString(),
kExpectedFakeCreds1DebugString);
std::shared_ptr<CallCredentials> creds2 =
GoogleIAMCredentials(kFakeToken2, kFakeSelector2);
context.set_credentials(creds2);
EXPECT_EQ(context.credentials(), creds2);
request.set_message("Hello");
request.mutable_param()->set_echo_metadata(true);
Status s = stub_->Echo(&context, request, &response);
EXPECT_TRUE(MetadataContains(context.GetServerTrailingMetadata(),
GRPC_IAM_AUTHORIZATION_TOKEN_METADATA_KEY,
kFakeToken2));
EXPECT_TRUE(MetadataContains(context.GetServerTrailingMetadata(),
GRPC_IAM_AUTHORITY_SELECTOR_METADATA_KEY,
kFakeSelector2));
EXPECT_FALSE(MetadataContains(context.GetServerTrailingMetadata(),
GRPC_IAM_AUTHORIZATION_TOKEN_METADATA_KEY,
kFakeToken1));
EXPECT_FALSE(MetadataContains(context.GetServerTrailingMetadata(),
GRPC_IAM_AUTHORITY_SELECTOR_METADATA_KEY,
kFakeSelector1));
EXPECT_EQ(context.credentials()->DebugString(),
kExpectedFakeCreds2DebugString);
EXPECT_EQ(request.message(), response.message());
EXPECT_TRUE(s.ok());
}
TEST_P(SecureEnd2endTest, AuthMetadataPluginKeyFailure) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
context.set_credentials(grpc::MetadataCredentialsFromPlugin(
std::unique_ptr<MetadataCredentialsPlugin>(
new TestMetadataCredentialsPlugin(
TestMetadataCredentialsPlugin::kBadMetadataKey,
"Does not matter, will fail the key is invalid.", false, true,
0))));
request.set_message("Hello");
Status s = stub_->Echo(&context, request, &response);
EXPECT_FALSE(s.ok());
6 years ago
EXPECT_EQ(s.error_code(), StatusCode::UNAVAILABLE);
EXPECT_EQ(context.credentials()->DebugString(),
kExpectedAuthMetadataPluginKeyFailureCredsDebugString);
}
TEST_P(SecureEnd2endTest, AuthMetadataPluginValueFailure) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
context.set_credentials(grpc::MetadataCredentialsFromPlugin(
std::unique_ptr<MetadataCredentialsPlugin>(
new TestMetadataCredentialsPlugin(
TestMetadataCredentialsPlugin::kGoodMetadataKey,
"With illegal \n value.", false, true, 0))));
request.set_message("Hello");
Status s = stub_->Echo(&context, request, &response);
EXPECT_FALSE(s.ok());
6 years ago
EXPECT_EQ(s.error_code(), StatusCode::UNAVAILABLE);
EXPECT_EQ(context.credentials()->DebugString(),
kExpectedAuthMetadataPluginValueFailureCredsDebugString);
}
TEST_P(SecureEnd2endTest, AuthMetadataPluginWithDeadline) {
ResetStub();
EchoRequest request;
request.mutable_param()->set_skip_cancelled_check(true);
EchoResponse response;
ClientContext context;
const int delay = 100;
std::chrono::system_clock::time_point deadline =
std::chrono::system_clock::now() + std::chrono::milliseconds(delay);
context.set_deadline(deadline);
context.set_credentials(grpc::MetadataCredentialsFromPlugin(
std::unique_ptr<MetadataCredentialsPlugin>(
new TestMetadataCredentialsPlugin("meta_key", "Does not matter", true,
true, delay))));
request.set_message("Hello");
Status s = stub_->Echo(&context, request, &response);
if (!s.ok()) {
6 years ago
EXPECT_TRUE(s.error_code() == StatusCode::DEADLINE_EXCEEDED ||
s.error_code() == StatusCode::UNAVAILABLE);
}
EXPECT_EQ(context.credentials()->DebugString(),
kExpectedAuthMetadataPluginWithDeadlineCredsDebugString);
}
TEST_P(SecureEnd2endTest, AuthMetadataPluginWithCancel) {
ResetStub();
EchoRequest request;
request.mutable_param()->set_skip_cancelled_check(true);
EchoResponse response;
ClientContext context;
const int delay = 100;
context.set_credentials(grpc::MetadataCredentialsFromPlugin(
std::unique_ptr<MetadataCredentialsPlugin>(
new TestMetadataCredentialsPlugin("meta_key", "Does not matter", true,
true, delay))));
request.set_message("Hello");
6 years ago
std::thread cancel_thread([&] {
gpr_sleep_until(gpr_time_add(gpr_now(GPR_CLOCK_REALTIME),
gpr_time_from_millis(delay, GPR_TIMESPAN)));
context.TryCancel();
});
Status s = stub_->Echo(&context, request, &response);
if (!s.ok()) {
6 years ago
EXPECT_TRUE(s.error_code() == StatusCode::CANCELLED ||
s.error_code() == StatusCode::UNAVAILABLE);
}
cancel_thread.join();
EXPECT_EQ(context.credentials()->DebugString(),
kExpectedAuthMetadataPluginWithDeadlineCredsDebugString);
}
TEST_P(SecureEnd2endTest, NonBlockingAuthMetadataPluginFailure) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
context.set_credentials(grpc::MetadataCredentialsFromPlugin(
std::unique_ptr<MetadataCredentialsPlugin>(
new TestMetadataCredentialsPlugin(
TestMetadataCredentialsPlugin::kGoodMetadataKey,
"Does not matter, will fail anyway (see 3rd param)", false, false,
0))));
request.set_message("Hello");
Status s = stub_->Echo(&context, request, &response);
EXPECT_FALSE(s.ok());
6 years ago
EXPECT_EQ(s.error_code(), StatusCode::UNAVAILABLE);
EXPECT_EQ(s.error_message(),
std::string("Getting metadata from plugin failed with error: ") +
kTestCredsPluginErrorMsg);
EXPECT_EQ(context.credentials()->DebugString(),
kExpectedNonBlockingAuthMetadataPluginFailureCredsDebugString);
}
TEST_P(SecureEnd2endTest, NonBlockingAuthMetadataPluginAndProcessorSuccess) {
auto* processor = new TestAuthMetadataProcessor(false);
StartServer(std::shared_ptr<AuthMetadataProcessor>(processor));
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
context.set_credentials(processor->GetCompatibleClientCreds());
request.set_message("Hello");
request.mutable_param()->set_echo_metadata(true);
request.mutable_param()->set_expected_client_identity(
TestAuthMetadataProcessor::kGoodGuy);
request.mutable_param()->set_expected_transport_security_type(
GetParam().credentials_type());
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(request.message(), response.message());
EXPECT_TRUE(s.ok());
// Metadata should have been consumed by the processor.
EXPECT_FALSE(MetadataContains(
context.GetServerTrailingMetadata(), GRPC_AUTHORIZATION_METADATA_KEY,
std::string("Bearer ") + TestAuthMetadataProcessor::kGoodGuy));
EXPECT_EQ(
context.credentials()->DebugString(),
kExpectedNonBlockingAuthMetadataPluginAndProcessorSuccessCredsDebugString);
}
TEST_P(SecureEnd2endTest, NonBlockingAuthMetadataPluginAndProcessorFailure) {
auto* processor = new TestAuthMetadataProcessor(false);
StartServer(std::shared_ptr<AuthMetadataProcessor>(processor));
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
context.set_credentials(processor->GetIncompatibleClientCreds());
request.set_message("Hello");
Status s = stub_->Echo(&context, request, &response);
EXPECT_FALSE(s.ok());
EXPECT_EQ(s.error_code(), StatusCode::UNAUTHENTICATED);
EXPECT_EQ(
context.credentials()->DebugString(),
kExpectedNonBlockingAuthMetadataPluginAndProcessorFailureCredsDebugString);
}
TEST_P(SecureEnd2endTest, BlockingAuthMetadataPluginFailure) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
context.set_credentials(grpc::MetadataCredentialsFromPlugin(
std::unique_ptr<MetadataCredentialsPlugin>(
new TestMetadataCredentialsPlugin(
TestMetadataCredentialsPlugin::kGoodMetadataKey,
"Does not matter, will fail anyway (see 3rd param)", true, false,
0))));
request.set_message("Hello");
Status s = stub_->Echo(&context, request, &response);
EXPECT_FALSE(s.ok());
6 years ago
EXPECT_EQ(s.error_code(), StatusCode::UNAVAILABLE);
EXPECT_EQ(s.error_message(),
std::string("Getting metadata from plugin failed with error: ") +
kTestCredsPluginErrorMsg);
EXPECT_EQ(context.credentials()->DebugString(),
kExpectedBlockingAuthMetadataPluginFailureCredsDebugString);
}
TEST_P(SecureEnd2endTest, CompositeCallCreds) {
ResetStub();
EchoRequest request;
EchoResponse response;
ClientContext context;
const char kMetadataKey1[] = "call-creds-key1";
const char kMetadataKey2[] = "call-creds-key2";
const char kMetadataVal1[] = "call-creds-val1";
const char kMetadataVal2[] = "call-creds-val2";
context.set_credentials(grpc::CompositeCallCredentials(
grpc::MetadataCredentialsFromPlugin(
std::unique_ptr<MetadataCredentialsPlugin>(
new TestMetadataCredentialsPlugin(kMetadataKey1, kMetadataVal1,
true, true, 0))),
grpc::MetadataCredentialsFromPlugin(
std::unique_ptr<MetadataCredentialsPlugin>(
new TestMetadataCredentialsPlugin(kMetadataKey2, kMetadataVal2,
true, true, 0)))));
request.set_message("Hello");
request.mutable_param()->set_echo_metadata(true);
Status s = stub_->Echo(&context, request, &response);
EXPECT_TRUE(s.ok());
EXPECT_TRUE(MetadataContains(context.GetServerTrailingMetadata(),
kMetadataKey1, kMetadataVal1));
EXPECT_TRUE(MetadataContains(context.GetServerTrailingMetadata(),
kMetadataKey2, kMetadataVal2));
EXPECT_EQ(context.credentials()->DebugString(),
kExpectedCompositeCallCredsDebugString);
}
TEST_P(SecureEnd2endTest, ClientAuthContext) {
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("Hello");
request.mutable_param()->set_check_auth_context(
GetParam().credentials_type() == kTlsCredentialsType);
request.mutable_param()->set_expected_transport_security_type(
GetParam().credentials_type());
ClientContext context;
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(response.message(), request.message());
EXPECT_TRUE(s.ok());
std::shared_ptr<const AuthContext> auth_ctx = context.auth_context();
std::vector<grpc::string_ref> tst =
auth_ctx->FindPropertyValues("transport_security_type");
ASSERT_EQ(1u, tst.size());
EXPECT_EQ(GetParam().credentials_type(), ToString(tst[0]));
if (GetParam().credentials_type() == kTlsCredentialsType) {
EXPECT_EQ("x509_subject_alternative_name",
auth_ctx->GetPeerIdentityPropertyName());
EXPECT_EQ(4u, auth_ctx->GetPeerIdentity().size());
EXPECT_EQ("*.test.google.fr", ToString(auth_ctx->GetPeerIdentity()[0]));
EXPECT_EQ("waterzooi.test.google.be",
ToString(auth_ctx->GetPeerIdentity()[1]));
EXPECT_EQ("*.test.youtube.com", ToString(auth_ctx->GetPeerIdentity()[2]));
EXPECT_EQ("192.168.1.3", ToString(auth_ctx->GetPeerIdentity()[3]));
}
}
class ResourceQuotaEnd2endTest : public End2endTest {
public:
ResourceQuotaEnd2endTest()
: server_resource_quota_("server_resource_quota") {}
void ConfigureServerBuilder(ServerBuilder* builder) override {
builder->SetResourceQuota(server_resource_quota_);
}
private:
ResourceQuota server_resource_quota_;
};
TEST_P(ResourceQuotaEnd2endTest, SimpleRequest) {
ResetStub();
EchoRequest request;
EchoResponse response;
request.set_message("Hello");
ClientContext context;
Status s = stub_->Echo(&context, request, &response);
EXPECT_EQ(response.message(), request.message());
EXPECT_TRUE(s.ok());
}
// TODO(vjpai): refactor arguments into a struct if it makes sense
std::vector<TestScenario> CreateTestScenarios(bool use_proxy,
bool test_insecure,
bool test_secure,
bool test_inproc,
bool test_callback_server) {
std::vector<TestScenario> scenarios;
std::vector<std::string> credentials_types;
grpc_core::ConfigVars::Overrides overrides;
overrides.client_channel_backup_poll_interval_ms =
kClientChannelBackupPollIntervalMs;
grpc_core::ConfigVars::SetOverrides(overrides);
#if TARGET_OS_IPHONE
// Workaround Apple CFStream bug
grpc_core::SetEnv("grpc_cfstream", "0");
#endif
if (test_secure) {
credentials_types =
GetCredentialsProvider()->GetSecureCredentialsTypeList();
}
auto insec_ok = [] {
// Only allow insecure credentials type when it is registered with the
// provider. User may create providers that do not have insecure.
return GetCredentialsProvider()->GetChannelCredentials(
kInsecureCredentialsType, nullptr) != nullptr;
};
if (test_insecure && insec_ok()) {
credentials_types.push_back(kInsecureCredentialsType);
}
// Test callback with inproc or if the event-engine allows it
CHECK(!credentials_types.empty());
for (const auto& cred : credentials_types) {
scenarios.emplace_back(false, false, false, cred, false);
scenarios.emplace_back(true, false, false, cred, false);
if (test_callback_server) {
// Note that these scenarios will be dynamically disabled if the event
// engine doesn't run in the background
scenarios.emplace_back(false, false, false, cred, true);
scenarios.emplace_back(true, false, false, cred, true);
}
if (use_proxy) {
scenarios.emplace_back(false, true, false, cred, false);
scenarios.emplace_back(true, true, false, cred, false);
}
}
if (test_inproc && insec_ok()) {
scenarios.emplace_back(false, false, true, kInsecureCredentialsType, false);
scenarios.emplace_back(true, false, true, kInsecureCredentialsType, false);
if (test_callback_server) {
scenarios.emplace_back(false, false, true, kInsecureCredentialsType,
true);
scenarios.emplace_back(true, false, true, kInsecureCredentialsType, true);
}
}
return scenarios;
}
INSTANTIATE_TEST_SUITE_P(
End2end, End2endTest,
::testing::ValuesIn(CreateTestScenarios(false, true, true, true, true)),
&TestScenario::Name);
INSTANTIATE_TEST_SUITE_P(
End2endServerTryCancel, End2endServerTryCancelTest,
::testing::ValuesIn(CreateTestScenarios(false, true, true, true, true)),
&TestScenario::Name);
INSTANTIATE_TEST_SUITE_P(
ProxyEnd2end, ProxyEnd2endTest,
::testing::ValuesIn(CreateTestScenarios(true, true, true, true, true)),
&TestScenario::Name);
INSTANTIATE_TEST_SUITE_P(
SecureEnd2end, SecureEnd2endTest,
::testing::ValuesIn(CreateTestScenarios(false, false, true, false, true)),
&TestScenario::Name);
INSTANTIATE_TEST_SUITE_P(
ResourceQuotaEnd2end, ResourceQuotaEnd2endTest,
::testing::ValuesIn(CreateTestScenarios(false, true, true, true, true)),
&TestScenario::Name);
} // namespace
} // namespace testing
10 years ago
} // namespace grpc
int main(int argc, char** argv) {
grpc::testing::TestEnvironment env(&argc, argv);
10 years ago
::testing::InitGoogleTest(&argc, argv);
int ret = RUN_ALL_TESTS();
return ret;
}