-
Notifications
You must be signed in to change notification settings - Fork 362
Fix target scaler fallback race condition on multi-worker plans #3190
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Closed
Closed
Changes from all commits
Commits
Show all changes
3 commits
Select commit
Hold shift + click to select a range
File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
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
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
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
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
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Outdated suggestions cannot be applied.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.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This method is all about taking metrics samples for incremental scaling. It doesn't make any sense to be making TBS calls here. Also, this approach makes a bunch of service calls whose results are just thrown away.
Instead, why don't we address the fundamental issue which is the _targetScalersInError collection as you're using it is supposed to be shared state across N workers, similar to metrics samples and snapshots. So one thing you could do is store that info in storage in the same way so its accessible by all instances.
Uh oh!
There was an error while loading. Please reload this page.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks @mathewc. A couple of points on the probe approach:
Service call frequency — the probe wasn't per-tick. It ran on first tick, then cached for 10 minutes (
_nextTargetScalerValidationTime). So it was one TBS call per scaler per 10 minutes, not "a bunch." For comparison, the concurrency snapshot (_concurrencyStatusRepository.ReadAsync) does a blob storage read on every SC call (~10s) with no caching.Storage-backed approach — I agree it's architecturally cleaner. I've implemented it in Fix target scaler fallback race condition via shared storage #3191 following the
BlobStorageConcurrencyStatusRepository/NullConcurrencyStatusRepositorypattern. It is a significantly bigger change (~700 lines vs ~150), but it fixes the root cause and keeps TBS calls out of the metrics sampling path.There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I do think I like the storage option better. Seems less hacky. One question I do have on this whole "fallback" approach - this ScaleMonitorService won't even be polling metrics if IsRuntimeScalingEnabled=false which is the normal case, right? Only if RTS is being used will IsRuntimeScalingEnabled=true. So usually TakeMetricsSamplesAsync won't even be running, so wondering how the fallback works in the normal case.