[Core] Fix the segfault from Opencensus upon shutdown #36906
Merged
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
Signed-off-by: SangBin Cho <[email protected]>
Signed-off-by: SangBin Cho <[email protected]>
Signed-off-by: SangBin Cho <[email protected]>
Signed-off-by: SangBin Cho <[email protected]>
rkooo567
commented
Jul 6, 2023
| @@ -146,7 +146,6 @@ CoreWorkerProcessImpl::CoreWorkerProcessImpl(const CoreWorkerOptions &options) | |||
|
|
|||
| CoreWorkerProcessImpl::~CoreWorkerProcessImpl() { | |||
| RAY_LOG(INFO) << "Destructing CoreWorkerProcessImpl. pid: " << getpid(); | |||
| RAY_LOG(DEBUG) << "Stats stop in core worker."; | |||
There was a problem hiding this comment.
we added the equivalent logs to stats::Shutdown
rickyyx
approved these changes
Jul 6, 2023
We are sending a request actually, and the problem is from protobuf, so it is unlikely related to Opencensus. It should be something shutdown ordering issue. I am not sure if my PR will actually fix the following issue, but client is not protected properly, and we should fix it anyway. |
rkooo567
added a commit
that referenced
this pull request
Jul 6, 2023
Seems like client_ in the metrics exporter is not lock-protected, but it is accessed by 2 thread (by a metrics export thread + task execution thread when ExportNow is called). I added a lock. I am not 100% sure if this will fix the issue because I was unable to reproduce the issue, but we can ask the user to try after merging this PR.
|
I am a bit concerned of the potential perf impact from the locks? |
|
@rickyyx the report code happens in a separate thread, and the main thread is only locked upon exit, so it should be okay. Let me know if you have other parts that could be problematic... |
|
@rkooo567 is there a cherry-pick PR? Thanks |
rkooo567
added a commit
to rkooo567/ray
that referenced
this pull request
Jul 11, 2023
) Seems like client_ in the metrics exporter is not lock-protected, but it is accessed by 2 thread (by a metrics export thread + task execution thread when ExportNow is called). I added a lock. I am not 100% sure if this will fix the issue because I was unable to reproduce the issue, but we can ask the user to try after merging this PR.
|
created here; #37311 |
SongGuyang
pushed a commit
to alipay/ant-ray
that referenced
this pull request
Jul 12, 2023
) Seems like client_ in the metrics exporter is not lock-protected, but it is accessed by 2 thread (by a metrics export thread + task execution thread when ExportNow is called). I added a lock. I am not 100% sure if this will fix the issue because I was unable to reproduce the issue, but we can ask the user to try after merging this PR. Signed-off-by: 久龙 <[email protected]>
bveeramani
pushed a commit
that referenced
this pull request
Jul 13, 2023
Seems like client_ in the metrics exporter is not lock-protected, but it is accessed by 2 thread (by a metrics export thread + task execution thread when ExportNow is called). I added a lock. I am not 100% sure if this will fix the issue because I was unable to reproduce the issue, but we can ask the user to try after merging this PR.
8 tasks
harborn
pushed a commit
to harborn/ray
that referenced
this pull request
Aug 17, 2023
) Seems like client_ in the metrics exporter is not lock-protected, but it is accessed by 2 thread (by a metrics export thread + task execution thread when ExportNow is called). I added a lock. I am not 100% sure if this will fix the issue because I was unable to reproduce the issue, but we can ask the user to try after merging this PR. Signed-off-by: harborn <[email protected]>
harborn
pushed a commit
to harborn/ray
that referenced
this pull request
Aug 17, 2023
) Seems like client_ in the metrics exporter is not lock-protected, but it is accessed by 2 thread (by a metrics export thread + task execution thread when ExportNow is called). I added a lock. I am not 100% sure if this will fix the issue because I was unable to reproduce the issue, but we can ask the user to try after merging this PR.
arvind-chandra
pushed a commit
to lmco/ray
that referenced
this pull request
Aug 31, 2023
) Seems like client_ in the metrics exporter is not lock-protected, but it is accessed by 2 thread (by a metrics export thread + task execution thread when ExportNow is called). I added a lock. I am not 100% sure if this will fix the issue because I was unable to reproduce the issue, but we can ask the user to try after merging this PR. Signed-off-by: e428265 <[email protected]>
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.
Why are these changes needed?
Seems like client_ in the metrics exporter is not lock-protected, but it is accessed by 2 thread (by a metrics export thread + task execution thread when ExportNow is called). I added a lock. I am not 100% sure if this will fix the issue because I was unable to reproduce the issue, but we can ask the user to try after merging this PR.
Related issue number
Fixes #36885
Checks
git commit -s) in this PR.scripts/format.shto lint the changes in this PR.method in Tune, I've added it in
doc/source/tune/api/under thecorresponding
.rstfile.