Description of problem:

Issue in file backup_and_restore/graceful-cluster-shutdown.adoc for shutting down the cluster gracefully, we have step 2 being cordoning the nodes and step 3 being evacuate all of the pods from the hosts.

In between these steps (2,3), we should call out that if you are running a software-defined storage solution like ODF on the cluster you need to first evict all workloads that are consuming storage first before evicting the rest of the cluster.

We have a KCS article that outlines this procedure with ODF: https://access.redhat.com/solutions/6642401

We could show some examples of how to remove workloads that are consuming ODF storage.

Additional info:

We have a doc we have created as well: https://hackmd.io/@johnsimcall/HJXwaZkJC