perf(benchmark): use grpc.SharedBufferPool for grpc.DialOption #718
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This was referenced Feb 25, 2024
This was referenced Feb 25, 2024
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #718 +/- ##
==========================================
+ Coverage 49.70% 49.75% +0.04%
==========================================
Files 235 235
Lines 20709 20715 +6
==========================================
+ Hits 10294 10306 +12
+ Misses 9773 9771 -2
+ Partials 642 638 -4 ☔ View full report in Codecov by Sentry. |
ijsong
force-pushed
the
feat_benchmark_gops_agent
branch
from
5c8e443 to
c387533
Compare
ijsong
force-pushed
the
feat_benchmark_gops_agent
branch
from
c387533 to
76484b8
Compare
ijsong
force-pushed
the
perf_grpc_client_sharedbuffer
branch
from
37baf23 to
889b990
Compare
ijsong
force-pushed
the
feat_benchmark_gops_agent
branch
from
76484b8 to
f370427
Compare
ijsong
force-pushed
the
perf_grpc_client_sharedbuffer
branch
from
889b990 to
01fbf39
Compare
ijsong
force-pushed
the
feat_benchmark_gops_agent
branch
from
f370427 to
9f31816
Compare
ijsong
force-pushed
the
perf_grpc_client_sharedbuffer
branch
2 times, most recently
from
c39f3e6 to
d604b6f
Compare
ijsong
changed the title
ijsong
force-pushed
the
perf_grpc_client_sharedbuffer
branch
from
d604b6f to
10a0efb
Compare
ijsong
changed the base branch from
main
to
bump_mockgen_from_0_3_0_to_0_4_0
ijsong
force-pushed
the
bump_mockgen_from_0_3_0_to_0_4_0
branch
from
6a06c96 to
5d97667
Compare
ijsong
force-pushed
the
perf_grpc_client_sharedbuffer
branch
from
10a0efb to
40506df
Compare
When a client library parses incoming gRPC messages, it allocates new buffers frequently. This can result in the creation of a large number of heap objects. However, gRPC offers an experimental feature called the shared buffer pool that can be used to address this issue. This pull request (PR) aims to optimize the gRPC receiving path using the shared buffer pool. Various experiments have shown that this optimization leads to a decrease in CPU time of runtime.gcBgMarkWorkers, depending on the workload. Since the shared buffer pool is experimental, it is applied only to the benchmark. Refs: - https://pkg.go.dev/google.golang.org/[email protected]/experimental#WithRecvBufferPool - https://pkg.go.dev/google.golang.org/[email protected]#SharedBufferPool
ijsong
force-pushed
the
perf_grpc_client_sharedbuffer
branch
from
40506df to
8014ece
Compare
hungryjang
approved these changes
Mar 5, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What this PR does
When a client library parses incoming gRPC messages, it allocates new buffers
frequently. This can result in the creation of a large number of heap objects.
However, gRPC offers an experimental feature called the shared buffer pool that
can be used to address this issue.
This pull request (PR) aims to optimize the gRPC receiving path using the shared
buffer pool. Various experiments have shown that this optimization leads to a
decrease in CPU time of runtime.gcBgMarkWorkers, depending on the workload.
Since the shared buffer pool is experimental, it is applied only to the
benchmark.
Refs: