It started a few days ago, and nothing really changed in our cluster. We just added a bunch of secrets, and noticed longer and longer deploys.
We have nothing fancy in the logs, and the only relevent event is :
Unable to mount volumes for pod "xxxxxx": timeout expired waiting for volumes to attach/mount for pod "xxxx"/"xxxx". list of unattached/unmounted volumes=[xxx-secrets yyyy-secrets zzzz-secrets ssl-certs xxxx-secrets default-token-n6pbo]
We have this event several times (it varies, let's say around 5 times), then the container starts as expected. It's an issue when it comes to single DB pod, the application is down for 5 minutes if the pod needs to restart.