configurable timeout for lifecycle/start

[berndocklin]

What is missing?

Make the timeout for startCassandra() REST call to management api lifecycle configurable.

Why do we need it?

In reconciling racks startCassandra() REST call has a very short time out of only 10 seconds before it kills the pods. The called lifecycle/start call was always synchronous, it has to wait until the start script fully executed and Cassandra/DSE starts can take much more time than that.

While it would be desirable to have this rest call async and then monitor state of starting-not-ready - currently that is not the case.

Environment

• Cass Operator version:

Needed in 1.10.x branches to be used in Astra Classic.

Anything else we need to know?: