Revise cluster management activity retry policies and timeouts.

[stoader]

Is your feature request related to a problem? Please describe.

Cluster create/update/delete operations may fail due an activity timing out. Often the timeout is caused by the cloud provider not fulfilling a requested operation within the the cadence activity's timeout, the operation on cloud provider side eventually completing successfully. In such scenarios an activity retry policy would help to continue the cluster operation workflow.

The total timeout of the activities that a workflow consists of may exceed the timeout set for the workflow leading to the workflow timing out prematurely.

Describe the solution you'd like to see Revise the retry policies and timeout settings of the cluster management workflows/activities and activity implementations to avoid the issues mentioned above.