Repository navigation
feat(operator): the chart runs two operator pods, on different nodes where it can - #673
Merged
Merged
Conversation
…where it can With one pod, a lost node holds every Kates resource until a new pod is scheduled and started. With two, the standby takes the Lease within 5 to 15 seconds. The pods prefer one node each unless affinity is set. CI's failure dump reads every operator pod's log, not one of them.
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 first step toward a mature operator: with `replicas: 1`, a lost node holds every `Kates` resource until a new pod is scheduled and started.
The controllers and the startup cleanup run only after a pod takes the Lease (`run.rs`), so the standby does nothing.
Proof
On a three-node Kind cluster, both pods ran on different nodes. With the Lease holder force-deleted, a new pod held the Lease 13 s later. In that run the freshly created replacement won the race, not the standby; on a real node loss the replacement can't start that fast. `helm lint` passes; `helm template --set replicas=1` renders no affinity.