Repair endpoints: repair related endpoints

[Miles-Garnsey]

This issue is a duplicate of one in our private mission control project.

To avoid the use of JMX we need to ensure that all of the current JMX endpoints that Reaper calls on Cassandra/DSE are available via HTTP instead.

The endpoints should be made available to management API, and then go on to call the Cassandra/DSE JMX functions below internally, without JMX traffic ever flowing over the network.

Please familiarise yourself with the design principles for the API https://github.com/riptano/mission-control/issues/492 prior to starting work. Note that all endpoints need to sit under /api/v2.

There is a list of endpoints here, this ticket deals only with those relating to repair resources.

Definition of done

Preview Give feedback

1. forceTerminateAllRepairSessions() DELETE /repairs
2. repairAsync() PUT /repairs with a JSON that represents the arguments String keyspace, Map<String, String> options returning a UUID corresponding to the repair.

[Miles-Garnsey]

This was closed via #358.