Mitigate memory leaks from long RequeueAfter periods #1989
Merged
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Outdated suggestions cannot be applied.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.
Fixes #1984.
We have an issue where the underlying timer used by client-go worker
queue implementation stays in memory until it expires.
Since one gets created at every reconciliation attempt, we end up with a
big bunch of timers in memory that will expire in 365 days by default.
To mitigate the memory leak, let's wait for no more than 10 hours to
reconcile. This does not prevent timers to leak, but restricts it to the next
10 hours.
This is done at the level of the aggregated results, to decouple this
workaround from any business logic like certs expiration. As opposed
to #1988 (closed).
The PR also refactors slightly the license controller to rely on the aggregated
Results structure. Hence benefit from the max 10 hour restriction.