Backport #29160 to v1.45.x: Switched to epoll1 as a default polling engine for Linux #29239
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
veblush
added
lang/core
release notes: yes
veblush
changed the title
markdroth
approved these changes
Mar 28, 2022
ctiller
approved these changes
Mar 28, 2022
lidizheng
added a commit
that referenced
this pull request
Apr 5, 2022
* Fix state machine (#29036) (#29055) * Bump version to 1.45.0-pre1 (on v1.45.x branch) (#29065) * bump version to 1.45.0-pre1 * regenerate projects * Remove GRPC_ENABLE_FORK_SUPPORT from setup.py (#29103) (#29120) * decrease verbosity of alts handshaker logs which can happen during cancellation (#29058) (#29119) * Include ADS stream error in XDS error updates (#29014) (#29121) * include ADS stream error in XDS error updates * Re-package XDS client errors into unavailable rather than unknown, in XDS resolver * TrySeq fix (#29112) (#29139) Co-authored-by: Craig Tiller <ctiller@google.com> * Bump version to 1.45.0 (on v1.45.x branch) (#29154) * bump version to 1.45.0 * regenerate projects * Fix C# "dotnet restore" on grpc-win2016 kokoro workers. (#29179) * avoid collision with pre-installed protoc on grpc-win2016 workers (#29180) * Switched to epoll1 as a default polling engine for Linux (#29239) * Revert "Remove GRPC_ENABLE_FORK_SUPPORT from setup.py (#29103) (#29120)" (#29230) This reverts commit 39dca57. * priority LB: avoid possibility of rescheduling a timer before it fires (#29188) (#29241) * priority LB: avoid possibility of rescheduling a timer before it fires * clang-format * fix memory leak * small change, just to be paranoid * inline StartFailoverTimerLocked() * initialize timer_pending_ to true * don't check shutting_down_ in timer callbacks Co-authored-by: Mark D. Roth <roth@google.com> * client channel: record call completion even if recv_trailing_metadata was not started (#29198) (#29242) * client channel: record call completion even if recv_trailing_metadata was not started * add test * add test for opencensus filter * remove unnecessary code Co-authored-by: Mark D. Roth <roth@google.com> * weighted_target_lb: avoid possibility of rescheduling a timer before it fires (#29203) (#29243) * weighted_target_lb: avoid possibility of rescheduling a timer before it fires * don't check shutdown_ or weight_ in timer callback * fix memory leak Co-authored-by: Mark D. Roth <roth@google.com> * Fix for cancelled calls (#29213) (#29240) Co-authored-by: Craig Tiller <ctiller@google.com> * xds_cluster_resolver: fix bug that caused us to incorrectly reuse priority child names (#29174) (#29244) * xds_cluster_resolver: fix bug that caused us to incorrectly reuse priority child names * add debug logging * assign child policy names within each discovery mechanism, not across all mechanisms * clang-format * add tests * don't need EDS service name or LOGICAL_DNS hostname in child name * remove unnecessary code * Decrease verbosity of log messages that can happen during ALTS handshake cancellation (#29255) (#29280) * Bump the version to v1.45.1 (#29278) * Version to 1.45.1 * Generate projects * priority LB: don't update picker while update is in flight (#29252) (#29314) Co-authored-by: Mark D. Roth <roth@google.com> * resolve conflict Co-authored-by: Craig Tiller <ctiller@google.com> Co-authored-by: AJ Heller <hork@google.com> Co-authored-by: Lidi Zheng <lidiz@google.com> Co-authored-by: apolcyn <apolcyn@google.com> Co-authored-by: Jan Tattermusch <jtattermusch@users.noreply.github.com> Co-authored-by: Esun Kim <veblush@google.com> Co-authored-by: Mark D. Roth <roth@google.com>
wanlin31
pushed a commit
that referenced
this pull request
Apr 19, 2022
* Fix state machine (#29036) (#29055) * Bump version to 1.45.0-pre1 (on v1.45.x branch) (#29065) * bump version to 1.45.0-pre1 * regenerate projects * Remove GRPC_ENABLE_FORK_SUPPORT from setup.py (#29103) (#29120) * decrease verbosity of alts handshaker logs which can happen during cancellation (#29058) (#29119) * Include ADS stream error in XDS error updates (#29014) (#29121) * include ADS stream error in XDS error updates * Re-package XDS client errors into unavailable rather than unknown, in XDS resolver * TrySeq fix (#29112) (#29139) Co-authored-by: Craig Tiller <ctiller@google.com> * Bump version to 1.45.0 (on v1.45.x branch) (#29154) * bump version to 1.45.0 * regenerate projects * Fix C# "dotnet restore" on grpc-win2016 kokoro workers. (#29179) * avoid collision with pre-installed protoc on grpc-win2016 workers (#29180) * Switched to epoll1 as a default polling engine for Linux (#29239) * Revert "Remove GRPC_ENABLE_FORK_SUPPORT from setup.py (#29103) (#29120)" (#29230) This reverts commit 39dca57. * priority LB: avoid possibility of rescheduling a timer before it fires (#29188) (#29241) * priority LB: avoid possibility of rescheduling a timer before it fires * clang-format * fix memory leak * small change, just to be paranoid * inline StartFailoverTimerLocked() * initialize timer_pending_ to true * don't check shutting_down_ in timer callbacks Co-authored-by: Mark D. Roth <roth@google.com> * client channel: record call completion even if recv_trailing_metadata was not started (#29198) (#29242) * client channel: record call completion even if recv_trailing_metadata was not started * add test * add test for opencensus filter * remove unnecessary code Co-authored-by: Mark D. Roth <roth@google.com> * weighted_target_lb: avoid possibility of rescheduling a timer before it fires (#29203) (#29243) * weighted_target_lb: avoid possibility of rescheduling a timer before it fires * don't check shutdown_ or weight_ in timer callback * fix memory leak Co-authored-by: Mark D. Roth <roth@google.com> * Fix for cancelled calls (#29213) (#29240) Co-authored-by: Craig Tiller <ctiller@google.com> * xds_cluster_resolver: fix bug that caused us to incorrectly reuse priority child names (#29174) (#29244) * xds_cluster_resolver: fix bug that caused us to incorrectly reuse priority child names * add debug logging * assign child policy names within each discovery mechanism, not across all mechanisms * clang-format * add tests * don't need EDS service name or LOGICAL_DNS hostname in child name * remove unnecessary code * Decrease verbosity of log messages that can happen during ALTS handshake cancellation (#29255) (#29280) * Bump the version to v1.45.1 (#29278) * Version to 1.45.1 * Generate projects * priority LB: don't update picker while update is in flight (#29252) (#29314) Co-authored-by: Mark D. Roth <roth@google.com> * resolve conflict Co-authored-by: Craig Tiller <ctiller@google.com> Co-authored-by: AJ Heller <hork@google.com> Co-authored-by: Lidi Zheng <lidiz@google.com> Co-authored-by: apolcyn <apolcyn@google.com> Co-authored-by: Jan Tattermusch <jtattermusch@users.noreply.github.com> Co-authored-by: Esun Kim <veblush@google.com> Co-authored-by: Mark D. Roth <roth@google.com>
wanlin31
pushed a commit
that referenced
this pull request
Apr 20, 2022
* Fix state machine (#29036) (#29055) * Bump version to 1.45.0-pre1 (on v1.45.x branch) (#29065) * bump version to 1.45.0-pre1 * regenerate projects * Remove GRPC_ENABLE_FORK_SUPPORT from setup.py (#29103) (#29120) * decrease verbosity of alts handshaker logs which can happen during cancellation (#29058) (#29119) * Include ADS stream error in XDS error updates (#29014) (#29121) * include ADS stream error in XDS error updates * Re-package XDS client errors into unavailable rather than unknown, in XDS resolver * TrySeq fix (#29112) (#29139) Co-authored-by: Craig Tiller <ctiller@google.com> * Bump version to 1.45.0 (on v1.45.x branch) (#29154) * bump version to 1.45.0 * regenerate projects * Fix C# "dotnet restore" on grpc-win2016 kokoro workers. (#29179) * avoid collision with pre-installed protoc on grpc-win2016 workers (#29180) * Switched to epoll1 as a default polling engine for Linux (#29239) * Revert "Remove GRPC_ENABLE_FORK_SUPPORT from setup.py (#29103) (#29120)" (#29230) This reverts commit 39dca57. * priority LB: avoid possibility of rescheduling a timer before it fires (#29188) (#29241) * priority LB: avoid possibility of rescheduling a timer before it fires * clang-format * fix memory leak * small change, just to be paranoid * inline StartFailoverTimerLocked() * initialize timer_pending_ to true * don't check shutting_down_ in timer callbacks Co-authored-by: Mark D. Roth <roth@google.com> * client channel: record call completion even if recv_trailing_metadata was not started (#29198) (#29242) * client channel: record call completion even if recv_trailing_metadata was not started * add test * add test for opencensus filter * remove unnecessary code Co-authored-by: Mark D. Roth <roth@google.com> * weighted_target_lb: avoid possibility of rescheduling a timer before it fires (#29203) (#29243) * weighted_target_lb: avoid possibility of rescheduling a timer before it fires * don't check shutdown_ or weight_ in timer callback * fix memory leak Co-authored-by: Mark D. Roth <roth@google.com> * Fix for cancelled calls (#29213) (#29240) Co-authored-by: Craig Tiller <ctiller@google.com> * xds_cluster_resolver: fix bug that caused us to incorrectly reuse priority child names (#29174) (#29244) * xds_cluster_resolver: fix bug that caused us to incorrectly reuse priority child names * add debug logging * assign child policy names within each discovery mechanism, not across all mechanisms * clang-format * add tests * don't need EDS service name or LOGICAL_DNS hostname in child name * remove unnecessary code * Decrease verbosity of log messages that can happen during ALTS handshake cancellation (#29255) (#29280) * Bump the version to v1.45.1 (#29278) * Version to 1.45.1 * Generate projects * priority LB: don't update picker while update is in flight (#29252) (#29314) Co-authored-by: Mark D. Roth <roth@google.com> * resolve conflict Co-authored-by: Craig Tiller <ctiller@google.com> Co-authored-by: AJ Heller <hork@google.com> Co-authored-by: Lidi Zheng <lidiz@google.com> Co-authored-by: apolcyn <apolcyn@google.com> Co-authored-by: Jan Tattermusch <jtattermusch@users.noreply.github.com> Co-authored-by: Esun Kim <veblush@google.com> Co-authored-by: Mark D. Roth <roth@google.com>
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.
This is a soft backport of #29160 removing epollex. This PR just changed the default polling engine to epoll1 while keeping epollex supported.