Closed Roming22 closed 5 months ago
Per discussion in slack, some of the upstream metrics, and all of the pipeline service exporter metrics, are accumulated after a pipelinerun is completed.
Odd timing issues, race conditions, could arise as if the deletes happen this fast.
Testing / more investigation is most likely needed. Possibly some sort of lessening of the time out, but not going immediately, might be needed.
We've a duplicate PR #868. I'll close that in favor of this one. However, as this comment on #868 mentions that we get into race conditions with immediate pruning, and there is a downstream JIRA issue dedicated to fixing it which is most likely needed to get this PR merged.
Do not merge this PR until the issue with logs not being synced before the resource is deleted is solved.
Do not merge this PR until the issue with logs not being synced before the resource is deleted is solved.
Add to ^^: until we do sufficient testing/verification that knative/controller-runtime won't aggregate events such that our various metrics, both upstream and the exporter, that are calculated after the pipelinerun is complete are not negatively impacted
@gabemontero @enarha I'll let you decide what you want to do with this PR. You may decide you do not want to delete PLR as soon as they complete.
@gabemontero @enarha I'll let you decide what you want to do with this PR. You may decide you do not want to delete PLR as soon as they complete.
yeah we are a ways from being able to do this IMO
rh-pre-commit.version: 2.0.3 rh-pre-commit.check-secrets: ENABLED