Improve safety of replace

[burmanm]

What is missing?
At the moment, it is possible for us to start replacing more nodes than the cluster can in reality handle. This is because we continue with the replace process of the next node before the first one has actually streamed all the data from other nodes.

The safety of this process is at the hands of the user, however a user-mistake could put the cluster in danger of losing availability (or worse).

Why do we need it?

Environment

• Cass Operator version:

  Insert image tag or Git SHA here

  **Anything else we need to know?**:

┆Issue is synchronized with this Jira Story by Unito
┆Issue Number: CASS-36