Open c4milo opened 3 months ago
I'm triple checking this one.
I deleted the Redpanda CR and deployed it from scratch and it is mounting correctly the volume now. I think the problem happens when those configurations are changed in a running cluster.
@c4milo is it possible for you to reproduce this issue? The chart itself doesn't seem to have anything that could result in the behavior you're describing.
The closest I've been able to reproduce is that it's possible to upgrade into a broken state depending on how you change the values file and that would stall out trying to update the highest ordinal Statefulset replica which could look like ordinal 1 had outdated volume mounts.
Did you see error logs from redpanda that would indicate that it was started up with an updated broker config but without the correct volume mounts?
I can try again, what I did was basically this:
Update: I have since walked around this by passing the certificates and keys directly through secret references.
What happened?
No custom provided
selfsigned
certs were mounted despite them being correctly configured inredpanda.yaml
. The change was done in a running cluster, I basically changed the CA from letsencrypt to self-signed. Provisioning an entirely new cluster works as expected.What did you expect to happen?
I was expecting it to mount the
selfsigned
certs into Redpanda's containers.How can we reproduce it (as minimally and precisely as possible)?. Please include values file.
Anything else we need to know?
Which are the affected charts?
Redpanda, Operator
Chart Version(s)
Cloud provider
JIRA Link: K8S-137