Repository navigation
feat(grounds-velocity): allow a recreate rollout via updateStrategy - #185
Merged
Merged
Conversation
The proxy always rolled out surge-first (maxSurge 1, maxUnavailable 0), so every image change ran a second Velocity JVM beside the old one. On a node with no headroom for that, the satellite rebooted mid-rollout. updateStrategy: Recreate starts the new pod only once the old one has drained and exited. maxSurge 0 would not achieve this, since a terminating pod does not count against it. RollingUpdate stays the default, so existing releases render unchanged.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What
Adds an
updateStrategyvalue togrounds-velocity:RollingUpdate(default)maxSurge: 1,maxUnavailable: 0Recreatestrategy.type: RecreateAny other value stops
helm templatewith a clear error.Why
The surge strategy was hard-coded, so every proxy image change started a second Velocity JVM next to the old one. On the stage satellite (cx33, no room to scale up because of the Hetzner limits), the node degraded and rebooted during the velocity-2 rollout to Velocity 4.2.0. Stage was down 16:11–16:28 UTC.
Recreatestarts the new pod only after the old one has drained and exited.maxSurge: 0would not do that: a terminating pod does not count against the surge, so the new pod would start next to a pod that is still draining. With drain enabled on stage, that overlap can last up to 15 minutes.Verified
helm template, default: same strategy block as before.helm template --set updateStrategy=Recreate:type: Recreatewith norollingUpdate(the API rejects that combination).--set updateStrategy=Bogus:updateStrategy must be RollingUpdate or Recreate, got "Bogus".drain.enabled=true,progressDeadlineSecondsis still rendered.helm lint: passes.Existing releases render unchanged.