Fix: ensure the uninstall wait to be properly done#4355
Conversation
build/Makefile
Outdated
| install: HELM_ARGS ?= | ||
| install: $(ensure-build-image) install-custom-pull-secret | ||
| -$(DOCKER_RUN) bash -c '[[ $$(helm status agones -n agones-system --output json | jq -r ".info.status") =~ (failed|pending-.*) ]] && helm uninstall agones --namespace=agones-system || true' | ||
| -$(DOCKER_RUN) bash -c '[[ $$(helm status agones -n agones-system --output json | jq -r ".info.status") =~ (failed|pending-.*) ]] && helm uninstall agones --namespace=agones-system --wait --timeout=10m || true' |
There was a problem hiding this comment.
(non blocking) Should we make the timeout 20m? 30m? Just in case
There was a problem hiding this comment.
I've put 10 mins to match the timeout from the install, but I don't mind increasing it to 20m in case there is something really blocking during the uninstall (it should uninstall the helm release only if it's in failed or pending)
Will let the pipeline finish and will push this change 👌🏼
|
Build Succeeded 🥳 Build Id: 4302d77d-9347-4248-af4f-d7b00ec8ca3c The following development artifacts have been built, and will exist for the next 30 days:
A preview of the website (the last 30 builds are retained): To install this version: |
|
Build Succeeded 🥳 Build Id: adb444cc-4f10-4b32-9032-0e37ecd5c344 The following development artifacts have been built, and will exist for the next 30 days:
A preview of the website (the last 30 builds are retained): To install this version: |
|
/gcbrun |
|
Build Succeeded 🥳 Build Id: dd590481-4b24-4d3b-985b-d3794133a13a The following development artifacts have been built, and will exist for the next 30 days:
A preview of the website (the last 30 builds are retained): To install this version: |
|
It's a pretty rare failure state. Might just be easier to merge and wait and see if it pops its head up again |
|
Build Succeeded 🥳 Build Id: 30a00156-dcad-44e4-a504-844a865ba1a2 The following development artifacts have been built, and will exist for the next 30 days:
A preview of the website (the last 30 builds are retained): To install this version: |
* fix: ensure the uninstall wait to be properly done * Fix: add more logs around previous invalid helm release state
What type of PR is this?
/kind bug
What this PR does / Why we need it:
Ensure we wait (with a timeout of 10m) for the helm uninstall when doing a helm upgrade --install, to ensure there is no race condition between the resources from the previous version installed and the new one
Which issue(s) this PR fixes:
Closes #4354
Special notes for your reviewer: