Yuxuan Li
|
98c0bd7c82
|
add setting channel args functionality to performance tesing
setting channel args --draft
clang-format
|
8 years ago |
Craig Tiller
|
e4fae4ef9f
|
clang-format
|
8 years ago |
Craig Tiller
|
f3ed2bf8ca
|
Fix interop
|
8 years ago |
yang-g
|
35674af901
|
make test robust
|
8 years ago |
yang-g
|
a3c95529c7
|
async works
|
8 years ago |
yang-g
|
8e708b12cb
|
WIP
|
8 years ago |
yang-g
|
50993b7a4c
|
async first take
|
8 years ago |
yang-g
|
9bef0740a4
|
Add test for explicit option
|
8 years ago |
yang-g
|
b74cf429fd
|
test pass
|
8 years ago |
yang-g
|
ad32764219
|
sync test pass
|
8 years ago |
Jan Tattermusch
|
9f89107f17
|
better qps_json_driver message when QPS_WORKERS env is missing
|
8 years ago |
Alexander Polcyn
|
4873d30ea2
|
allow disable core_list setting and override qps server in benchmarks
|
8 years ago |
David Garcia Quintas
|
855a1063ec
|
grpclb: skip slash in server uri path if present
|
8 years ago |
Sree Kuchibhotla
|
7eef316e20
|
Fix metrics server
|
8 years ago |
Yuchen Zeng
|
d37f642f35
|
Support server streaming
Skip unparsable input
Add tests for uni-directional stream calls
Simplify client stream handling
|
8 years ago |
Yuchen Zeng
|
f9329217b1
|
Support client streaming
|
8 years ago |
Yuchen Zeng
|
a84cdb8c81
|
Add parse, tobinary, totext commands
|
8 years ago |
Mark D. Roth
|
13f3574642
|
clang-format
|
8 years ago |
Ronak Jain
|
a3f8097b34
|
replaced protobuf tag in compile_test_golden
|
8 years ago |
Mark D. Roth
|
473ff83fac
|
Fix filter_end2end_test.
|
8 years ago |
Mark D. Roth
|
42663fb20e
|
Fix bug.
|
8 years ago |
Mark D. Roth
|
aa1cd14729
|
Clean up C++ filter API.
|
8 years ago |
Craig Tiller
|
9e90d22ab7
|
Fixes
|
8 years ago |
yang-g
|
6c8de363ca
|
Use user provided method name when using binary files in cli
|
8 years ago |
Craig Tiller
|
f658bf0e1b
|
Start resolving memory issues in C++ metadata
|
8 years ago |
Craig Tiller
|
18348a3bfb
|
Fix C++ test compilation
|
8 years ago |
David Garcia Quintas
|
ff3112188b
|
Renamed google_benchmark submodule to benchmark
|
8 years ago |
Craig Tiller
|
25910a2896
|
Fix compilation
|
8 years ago |
Alex Polcyn
|
fcf09ea42e
|
handle empty string for qps workers in driver and dont quit them on netperf
|
8 years ago |
Craig Tiller
|
76b7ab431a
|
Interop test might take a little longer to start up (esp. w/tsan)
|
8 years ago |
Mark D. Roth
|
bf25e240bc
|
Fix test name conflict in import. Also use low-thread-count variant for asan.
|
8 years ago |
Craig Tiller
|
fbcd9a7344
|
Update server_crash_test to account for longer reconnect timeouts
|
8 years ago |
Mark D. Roth
|
d101e14211
|
Revert "Renamed google_benchmark submodule to benchmark"
|
8 years ago |
David Garcia Quintas
|
7fa08e24b0
|
Fix data race with atomic for interop c++
|
8 years ago |
David Garcia Quintas
|
3c5a868a1b
|
PR comments #2
|
8 years ago |
David Garcia Quintas
|
523a4aa0a0
|
PR comments
|
8 years ago |
David Garcia Quintas
|
033c51e8c8
|
Deflake grpclb test
|
8 years ago |
David Garcia Quintas
|
d6936b6bee
|
PR comments
|
8 years ago |
Craig Tiller
|
3b45b8d60b
|
Revert "Test credentials provider update"
|
8 years ago |
Craig Tiller
|
9564cf9f0a
|
Take into account all the configs
|
8 years ago |
Craig Tiller
|
d04c822b58
|
Unique-ify uds address between concurrent bm_fullstack runs
|
8 years ago |
Craig Tiller
|
e89e54f4b6
|
clang-format code
|
8 years ago |
Mark D. Roth
|
c56d3ebd7e
|
Improve logging.
|
8 years ago |
David Garcia Quintas
|
6deadf5b88
|
Propagate termination signals to subprocesses
|
8 years ago |
Craig Tiller
|
c991e80e07
|
Limit outstanding rpc count for synchronous {client,server} tsan tests
|
8 years ago |
David Garcia Quintas
|
d449003fe6
|
Shutdown grpc for fullstack benchmark
|
8 years ago |
Craig Tiller
|
ab34b12e50
|
Make a better educated guess about CPU costs of a benchmark
|
8 years ago |
Craig Tiller
|
f739c058b7
|
Make a better educated guess about CPU costs of a benchmark
|
8 years ago |
Mark D. Roth
|
37d5f9b4e5
|
Add debug logging for process termination state.
|
8 years ago |
Craig Tiller
|
6d74225615
|
Reduce pre-generated key count
Doesn't affect benchmark accuracy, but makes startup time substantially
faster, and hopefully eliminates occurences like this:
https://grpc-testing.appspot.com/job/gRPC_master_linux/642/testReport/junit/(root)/c++_linux_dbg/bins_dbg_bm_fullstack_GRPC_POLL_STRATEGY_epoll/
|
8 years ago |