SOLR-18284: Load-average and memory circuit breakers no longer stampede or trip on transient pre-GC heap peaks#4514
Open
markrmiller wants to merge 1 commit into
Open
Conversation
…de or trip on transient pre-GC heap peaks Load-average and memory circuit breakers were doing more harm than good under high concurrency. LoadAverageCircuitBreaker called OperatingSystemMXBean.getSystemLoadAverage() on every request. The OS load average is a one-minute moving average, so re-polling it per-request is wasted work - and when many requests arrived concurrently, the syscall stampede hammered the CPU, which is the condition this breaker exists to prevent rather than cause. MemoryCircuitBreaker sampled MemoryMXBean.getHeapMemoryUsage().getUsed() on a 30-second moving average. With a generational collector that signal climbs toward max between collections during normal operation; the breaker would trip on transient pre-GC peaks that GC was about to reclaim.
github-actions
Bot
added
documentation
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Load-average and memory circuit breakers were doing more harm than good under high concurrency.
LoadAverageCircuitBreaker called OperatingSystemMXBean.getSystemLoadAverage() on every request. The OS load average is a one-minute moving average, so re-polling it per-request is wasted work - and when many requests arrived concurrently, the syscall stampede hammered the CPU, which is the condition this breaker exists to prevent rather than cause.
MemoryCircuitBreaker sampled MemoryMXBean.getHeapMemoryUsage().getUsed() on a 30-second moving average. With a generational collector that signal climbs toward max between collections during normal operation; the breaker would trip on transient pre-GC peaks that GC was about to reclaim.
https://issues.apache.org/jira/browse/SOLR-18284