-
Notifications
You must be signed in to change notification settings - Fork 249
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
Compact reports deleting more rows than revisions #363
Comments
Also - with the change from #360 the inner loop will keep moving the target ahead if additional revisions are created while it's running. Note the last few iterations only compact 1 or 2 revisions:
|
Confirmed that the additional deleted rows are from resources deleted by TTL. |
Merged
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
After running a load test against the cluster, compaction reports deleting more rows than revisions. I'm not sure if this is a bug or what, but it merits investigation.
Note that it does eventually settle out after some, a while after the load test is done. Perhaps related to rows with TTLs expiring? These would be deleted directly from the datastore, without a new resource version being written with a deletion timestamp.
kine/pkg/logstructured/logstructured.go
Lines 332 to 351 in 7966aec
We should probably have some metrics around workqueue depth and delay to see if massive loading of events (one of the things that the load test does) is pushing the TTL workqueue harder than it currently handles well.
The text was updated successfully, but these errors were encountered: