Would the latency of a persistent volume claim on Kubenetes make my service unavailable?

Question

Within Why does a GCE volume mount in a kubernetes pod cause a delay? is:

But when the pod has a GCE persistent volume mount, the Running state is reached somewhere between 20 and 60 seconds.

So for example when doing a rolling deployment, does this mean there is a 20-60 seconds of the service being unavailable?

Answer 1

When you do a rolling update, pods will not be teared down all at once. So your service will not interrupt. And that question is pretty old, k8s have many improvement ever since. In my case(GKE with latest k8s), the (re)start delay is very short.