-
Notifications
You must be signed in to change notification settings - Fork 9
Description
Is there an existing issue already for this bug?
- I have searched for an existing issue, and could not find anything. I believe this is a new bug.
I have read the troubleshooting guide
- I have read the troubleshooting guide and I think this is a new bug.
What happened?
We built a plugin to add a sidecar container to the cluster pod. The plugin seems to work, when we add it when we create a new cluster object.
However, when we update an existing cluster (to add the plugin configuration), then no lifecycles are triggered.
We currently listen to the following:
pod: create, update, evaluate- role: create, update, evaluate
We also briefly tried the operator implementation (mutate cluster), but it's unclear if that is an appropriate measure in this case? It's also a bit complicated to get right. I don't want to rule out that we are doing anything wrong either. :D
Does anyone have any pointers which hooks we should use?
Another idea (to quickfix) is to add an annotation to the cluster object to hopefully propagate to all the relevant kubernetes objects (pod, role) in order to then trigger the lifecycle hooks.
But that also doesn't seem right? Any guidance is appreciated.
Thank you!
Cluster resource
Relevant log output
Code of Conduct
- I agree to follow this project's Code of Conduct