* Check if memory owner available prior to polling it
The transport may drop the memory owner during its destruction sequence
* tcp_fix
* Revert "Revert "New resource quota integration (#27643)" (#28014)"
This reverts commit 0ea2c37263.
* clang-format
* fix-path
* fix
The authz test flaked as no RPCs of the expected type had completed
within the sampling window. Server logs showed authz logs completing
batch of 276 RPCs back-to-back, without the expected 40 ms separation
(qps=25). It took a bit over 1 second to process through the backlog.
With the sample duration of 500 ms and there being a polling delay
between when the channel is READY and when the test driver polls
channelz, it makes sense that we can get lucky much of the time.
Obviously, adding a sleep isn't great either, but measuring the queue
length indirectly is more complex than really appropriate here. The real
solution is to stop using this continuous-qps test client.
```
Traceback (most recent call last):
File "/tmp/work/grpc/tools/run_tests/xds_k8s_test_driver/tests/authz_test.py", line 252, in test_tls_allow
grpc.StatusCode.OK)
File "/tmp/work/grpc/tools/run_tests/xds_k8s_test_driver/tests/authz_test.py", line 183, in configure_and_assert
method=rpc_type)
File "/tmp/work/grpc/tools/run_tests/xds_k8s_test_driver/framework/xds_k8s_testcase.py", line 284, in assertRpcStatusCodes
self.assertGreater(stats.result[status_code.value[0]], 0)
AssertionError: 0 not greater than 0
```
* Reintroduce the EventEngine default factory
An application can provide an EventEngine factory function that allows
gRPC internals to create EventEngines as needed. This factory would be
used when no EventEngine is provided for some given channel or server,
and where an EventEngine otherwise could not be provided by the
application. Note that there currently is no API to provide an
EventEngine per channel or per server.
I've also deleted some previous iterations on global EventEngine and
EventEngine factory ideas. This new code lives in a public API, and
coexists with iomgr instead of being isolated to an EventEngine-specific
iomgr implementation.
* add proper namespaces, and fix description
* put factory functions in their own file (for replaceability)
* add synchronization
* generate_projects.sh
* extract event_engine_base and event_engine_factory targets
Also separate iomgr/event_engine files in the BUILD, with comments
* gpr_platform
* move all EE factory declarations to event_engine_base
Makes internal hackery easier.
* add missing deps
* reorder dep alphabetically
* comment style change
* Run 2to3 on tools directory
* Delete github_stats_tracking
* Re-run 2to3
* Remove unused script
* Remove unused script
* Remove unused line count utility
* Yapf. Isort
* Remove accidentally included file
* Migrate tools/distrib directory to python 3
* Remove unnecessary shebang
* Restore line_count directory
* Immediately convert subprocess.check_output output to string
* Take care of Python 2 shebangs
* Invoke scripts using a Python 3 interpreter
* Yapf. Isort
* Try installing Python 3 first
* See if we have any Python 3 versions installed
* Add Python 3.7 to Windows path
* Try adding a symlink
* Try to symlink differently
* Install six for Python 3
* Run run_interop_tests with python 3
* Try installing six in python3.7 explicitly
* Revert "Try installing six in python3.7 explicitly"
This reverts commit 2cf60d72f3.
* And debug some more
* Fix issue with jobset.py
* Add debug for CI failure
* Revert microbenchmark changes
* Add back references and scope field
* Set scope in router
* Reverse order of cleanup
* Add router_scope flag
* Use router_scope flag to create Router
* I apparently don't know how to brain
* Yapf
* Yeah, that can't be the default
* Remove debug print
* Remove impossible todos
* And another
* Switch from router-scope to config-scope
* Implement schema changes
* Use backend service URL
* Use CLH reference format to backend service
* I am an idiot
* *internal screaming*
* Try project number
* Why is this all awful
* Go back to trying project name
* Try cleaning things up
* Agh
* Address review comments
* Remove superfluous Optional type
A broken fix for the server-side bug is producing invalid configuration,
causing the server to reject all the configuration. Disable the
configuration and tests until fix is fixed.
* Fix all lint errors in repo.
* Use strict buildifier by default
* Whoops. That file does not exist
* Attempt fix to buildifier invocation
* Add missing copyright
* commit after sync
* Moving to v1.41.1
* Attempt to fix interop test
* Increase test timeout
* Regenerate testcases. Fix old Python images to 3.7
Co-authored-by: Donna Dionne <donnadionne@google.com>
* make it obvious which gke benchmark script is experimental
* add cfg file for new kokoro jobs with a more fitting name
* update performance readme after renaming _v2 benchmark job
* fix typo
* Upgrade benchmark to 1.6.0 and remove hacks.
Details:
- GRPC currently uses an old version of benchmark (from Sept 2020). It should probably upgrade because downstream, in google3, everyone is already using 1.6.0)
- Removed the hack added in PR/27629 to allow benchmarks in GRPC to continue to work with both pre-1.6.0 and 1.6.0 benchmarks.
(This was needed to allow importing benchmarks 1.6.0 into google3 without breaking GRPC)
* fix typo
* update third_party/benchmark and check_submodules.sh
* Upmerge from v1.41.x (#27821)
* Bump version to v1.41.0-pre1 (#27371)
* Bump version to v1.41.0-pre1
* Regenerate projects
* [Backport #27373] add testing_version flag (#27385)
* Bump version to v1.41.0-pre2 (#27390)
* Bump version to v1.41.0-pre2
* Regenerate projects
* Core 19: bump core version from 18.0.0 to 19.0.0 (#27394)
* Bump core version to 19.0.0
* Regenerate projects
* fix use-after-free metadata corruption in C# when receiving response headers for streaming response calls (#27398)
* Final release: bump up version to 1.41.0 (#27476)
* Bump version to 1.41.0
* Regenerate projects
* xds_k8s_test: increase timeout to 3 hours due to recent timeout failure (#27580)
* Revert "xds_k8s_test: increase timeout to 3 hours due to recent timeout failure (#27580)" (#27590)
This reverts commit da0c7d680f.
* Update root pem certs (backport of #27539) (#27619)
* Update boringssl to the latest (#27606) (#27625)
* Change boringssl branch name
* update submodule boringssl-with-bazel with origin/main-with-bazel
* update boringssl dependency to main-with-bazel commit SHA
* regenerate files
* Increment podspec version
* generate boringssl prefix headers
* Bumping up version to v1.41.1 (#27699)
* Bump version to v1.41.1
* Regenerate projects
* [Backport][v1.41.x] xds-k8s tests: Use test driver from master branch (#27695)
Backports sourcing the test driver install script from master.
This is a backport of #27389, #27462 and #27658:
* Add missing quatation marks.
These were missed when creating the Python virtual env.
* xds-k8s tests: Use test driver from master branch (#27462)
Instead of directly sourcing the test driver provisioning script from the same branch, the script is downloaded (with curl) and sourced from the master branch.
This allows changes made to the test driver to be reflected in all future release branches. A separate PR will backport this change to existing release branches.
All cluster definitions are also moved to the install script, allowing any cluster changes to be done in one place in the master branch.
* xds_k8s tests: Fix xlang install script sourcing. (#27658)
This change sources the test driver install script correctly for the xlang tests.
This fixes a mistake in #27462 where this was missed.
* Fix Python Interop (#27620) (#27703)
* WIP. Attempt to fix interop
* Yapf
* Switch Python xDS Example Server to Listen on IPV4 Only (#27679)
* Switch to IPV4
* Update to all hosts
* Fix rvm ruby install failure (#27769)
Co-authored-by: donnadionne <donnadionne@google.com>
Co-authored-by: Lidi Zheng <lidiz@google.com>
Co-authored-by: Jan Tattermusch <jtattermusch@users.noreply.github.com>
Co-authored-by: sanjaypujare <sanjaypujare@users.noreply.github.com>
Co-authored-by: Sergii Tkachenko <sergiitk@google.com>
Co-authored-by: Esun Kim <veblush@google.com>
Co-authored-by: Terry Wilson <terrymwilson@gmail.com>
Co-authored-by: Richard Belleville <rbellevi@google.com>
* added perf_counters.cc manually since the script didn't work
Co-authored-by: Mark D. Roth <roth@google.com>
Co-authored-by: donnadionne <donnadionne@google.com>
Co-authored-by: Lidi Zheng <lidiz@google.com>
Co-authored-by: Jan Tattermusch <jtattermusch@users.noreply.github.com>
Co-authored-by: sanjaypujare <sanjaypujare@users.noreply.github.com>
Co-authored-by: Sergii Tkachenko <sergiitk@google.com>
Co-authored-by: Esun Kim <veblush@google.com>
Co-authored-by: Terry Wilson <terrymwilson@gmail.com>
Co-authored-by: Richard Belleville <rbellevi@google.com>
* Support generic_xds_config in GCE/GKE frameworks and Python tests
* Correct the field name
* Address comments
* Add new CSDS support to the regular K8s testcases
* Support incomplete CSDS config dump
The control plane was updated to more properly match the principal being
present, so now plaintext and mTLS are working properly. But the change
is using slightly the wrong semantics for TLS, so we get to change which
tests are commented out.
This commit is part of the effort to create binder channel with
GRPC_CLIENT_CHANNEL type.
The resolver will be used during name resolution, and the result will
later be used to identify the corresponding endpoint binder in
SubchannelConnector.
Besides the unit test, this change is tested with other changes locally
end to end on real device.
The pool serves as a buffer for interaction between C++ and Java.
The buffer can let us avoid calling into Java code in channel connector
implementation (which has not been merged yet), simplifies interaction
between C++ and Java.
Temporary changes are made to channel_create.cc to keep the example apps
working but they will be rewrite after we start creating client channel
instead of direct channel.
* RLS LB policy plugin implementation
* Add RLS tests
* rls proto
* generate projects
* Fix adding child policy to child policy map
* Fix issues
* Fix rotten bits
* generate projects
* revert some strange changes
* First batch of fixes
* second batch of addressing comments
* generate_projects
* 3rd batch of fixes
* generate_projects
* 4th batch of fix
* 5th batch of fix
* Empty commit to trigger github PR refresh
* Add insecure build
* generate project
* update picker in a closure
* mutex guard audit
* clang-format
* Aggregate child policy states
* Use OrphanablePtr on ChildPolicyOwner()
* More fixes on comments
* Remove include grpc_security from rls.cc
* Fix key builder test as we removed RlsFindPathFromMetadata
* Update rls proto and add multi-target capability
* code changes to fix build after merging master
* remove support for insecure builds
* WIP
* add LB policy metadata lookup API
* add API for creating errors from C++ strings
* add missing include
* finished updating JSON parsing
* use DualRefCounted<> for ChildPolicyWrapper
* use grpc_core::Mutex instead of std::recursive_mutex
* add lock annotations and fix some lifetime issues
* misc fixes and cleanups
* simplify child policy connectivity state machine
* add comment about hopping into ExecCtx
* don't use wait_for_ready for RLS call
* extraKeys and constantKeys support
* add lock annotations for child policy wrapper connectivity state
* improve logging
* simplify child policy state handling in picker
* use C++ style comments
* remove rls_config.proto, since it's not needed
* rewrote tests
* fix build
* fix copyright headers
* clang-format
* appease clang-tidy
* remove unnecessary dependendency on grpclb
* buildifier
* remove illegal term
* remove unneeded includes
* clang-format
* fix clang-tidy
* fix build
* move class declaration into .cc file
* clang-tidy again
* fix build
* reorganize code and misc cleanups
* clang-format
* fixed a bunch of asan bugs; still some left to debug
* fix asan problem
* make test work on IPv6-only machines
* move LRU list handling into Cache::Entry and add lock annotations
* use preincrement instead of postincrement
* fix deadlock
* add more FIXMEs
* clean up backoff timer logic
* fix build from merge
* clang-format
* fix include path to work on podspec builds
* clean up picker logic
* clang-format
* update rls.proto
* populate reason and stale_header_data in RLS request
* improve logging and fix some cache size type issues
* clang-format
* add some TODOs about a better way to expose channel creds to LB policies
* centralize handling of channel's default authority
* clang-format
* fix backup poller bug
* fix handling of call creds and authority on RLS channel
* remove unused params
* clang-format
* use two-phase update for child policies to avoid deadlocks
* fix portability issue
* minor cleanups
* update for change in grpc_error_get_status() API
* change test to store RLS requests and responses in proto form
* clang-format
* account for test slowdown factor in timeouts
* fix tsan failure and channelz linkage
* fix RLS authority death test
* fix test to not try to connect to target that should not work
* remove illegal term
* cacheSizeBytes is a required field
* add missing BUILD deps from merge
* add tests for connectivity state reporting and fix bugs found
* fix BUILD sanity
* buildifier
* fix BUILD package path
* fix bugs related to child policy lifetime and updates
* remove unnecessary srand() call from grpc_init()
* add test for two cache entries with the same target
* update rls.proto
* change RLS config parser test to support GRPC_ERROR_IS_ABSEIL_STATUS
* update upb codegen
* fix include path
* add env var guard
* avoid duplication of CountedService
* generate_projects
* add upper limit for cacheSizeBytes
* fix build
Co-authored-by: Muxi Yan <mxyan@google.com>
Craig Tiller
Lidi Zheng
ZhenLian
Eric Anderson
Richard Belleville
AJ Heller
Paulo Castello da Costa
brandonpaiz
Denny C. Dai
Jan Tattermusch
Hannah Shi
Vy Nguyen
yifeizhuang
Yash Tibrewal
Ming-Chuan
Mark D. Roth