feat: mute collectors based on exception count #56
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
ansjcy
reviewed
Nov 13, 2023
| // since other collectors run every 5 sec, we can obtain an estimate | ||
| // of the exceptions they've thrown in the last ~10 runs or so. | ||
| int exceptionCount = StatsCollector.instance().exceptionCount(collector.getErrorMetric()); | ||
| if (!collector.inProgress() || exceptionCount > 5) { |
There was a problem hiding this comment.
We can do exceptionCount > 5 in the POC, but in the final implementation, this threshold should be configurable.
ansjcy
reviewed
Nov 13, 2023
| @@ -21,9 +21,7 @@ | |||
| */ | |||
| public class ServiceMetrics { | |||
| public static SampleAggregator READER_METRICS_AGGREGATOR, | |||
| STAT_METRICS_AGGREGATOR = new SampleAggregator(StatMetrics.values()), | |||
There was a problem hiding this comment.
Any specific reasons we removed those aggregators? Are those just for testing?
There was a problem hiding this comment.
These aggregators are unused in the entire Performance Analyzer codebase. They seem to be leftover from an earlier refactoring possibly involving RCA.
Do you think this is a change that would benefit from being its own PR?
and use them in deciding whether to mute a collector
caused by Mockito ref: https://stackoverflow.com/questions/68113065/mockito-inaccessible-object-exception-while-creating-a-mock-object To quote: "This can happen if Mockito requires reflective access to non-public parts in a Java module. you can get around this by explicitly allowing access via --add-opens in your java call"
invokePrivilegedAndLogError already exists for that purpose consider transitioning current uses to invokePrivilegedAndLogError this is important as the exception counting mechanism relies on exceptions not being swallowed earlier in call stack
by adding repetition options to runBehavior
kartva
force-pushed
the
exception-count-muting
branch
from
89de19a
to
7658b58
Compare
will file separate PR
|
What sort of configuration setting do you think would be appropriate for |
|
summary of changes so far:
|
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.
Is your feature request related to a problem? Please provide an existing Issue # , or describe.
As discussed at opensearch-project/performance-analyzer#549
Describe the solution you are proposing
We use the exception counts maintained by
StatsCollectorto dynamically mute collectors based on the number of exceptions they throw in the last minute of sampling.Additional context
Currently in the proof of concept stage. If this is successful, further avenues of investigation may be:
Check List
By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.
For more information on following Developer Certificate of Origin and signing off your commits, please check here.