Set node drainpolicy for longhorn storage class `single-replica-storageclass`

What does this MR do and why?

We are using longhorn storage class called single-replica-storageclass which uses single replica.

By default, longhorn uses block-if-contains-last-replica node drain policy.

This will block the node drain during rolling upgrade.

This MR fixes such issues which is also described at #1160 (comment 1862225563)

Related reference(s)

https://longhorn.io/docs/1.5.4/references/settings/#node-drain-policy

Test coverage