goschedstats: add cluster setting to always do short sampling #133459
+98
−6
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
The adaptive sampling, with the long period equal to 250ms, can result in sluggish changes in the number of AC slots, resulting in unnecessary queueing. The 1ms sampling is cheap, even on an idle roachprod node. Fixes cockroachdb#131766 Epic: none Release note (ops change): The goschedstats.always_use_short_sample_period.enabled setting should be set to true for any serious production cluster, to prevent unnecessary queuing in admission control CPU queues.
|
This change is |
sumeerbhola
added
backport-23.2.x
aadityasondhi
approved these changes
Oct 25, 2024
There was a problem hiding this comment.
Reviewable status:
complete! 1 of 0 LGTMs obtained (waiting on @RaduBerinde and @sumeerbhola)
pkg/util/goschedstats/runnable.go line 88 at r1 (raw file):
"goschedstats.always_use_short_sample_period.enabled", "when set to true, the system always does 1ms sampling of runnable queue lengths", false)
Any reason why we don't just make this true by default. considering the CPU profile helps show that this is cheap to do?
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Outdated suggestions cannot be applied.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.
The adaptive sampling, with the long period equal to 250ms, can result in sluggish changes in the number of AC slots, resulting in unnecessary queueing. The 1ms sampling is cheap, even on an idle roachprod node.
Effect of enabling this on an idle node -- the CPU utilization does not change
A 15s cpu profile on this 8vCPU node shows only 20ms in goschedstats and callbacks.
Fixes #131766
Epic: none
Release note (ops change): The
goschedstats.always_use_short_sample_period.enabled setting should be set to true for any serious production cluster, to prevent unnecessary queuing in admission control CPU queues.