See issue #6079 as the roll-up for node upgrades.

Use case: we upgrade nodes by tearing them down (deleting the instances) and bringing up new ones. Do data on the nodes survive?

This might go beyond just 'volumes' if there are other ways that data should persist on Kubernetes—anything that might rely on the presence of a node.

This issue will contain anywhere from zero to a significant amount of work to ensure that storage is migrated safely, and as a stretch goal, smartly.

References:

• The most relevant discussion regarding storage and upgrades is in upgrade: Make node components upgradeable (at all possible) #3333.
• This is ostensibly part of smartly evacuating nodes before upgrading them Node upgrades: graceful node evacuation #6080