Create and expose metrics for 1-1 restore service

[karol-kokoszka]

As 1-1 restore is going to be a new service in Scylla Manager, there is a need to create new set of metrics.

There must be a possibility to:

• track the SSTables copy part
  First, SM must count how much of the data must be copied (size is a part of manifest file). Then, SM must periodically query RClone for the given job-id to see what amount of bytes is copied already.
  scylla_manager_<clone_restore_or_other_name>_remaining_bytes{source_cluster_id, cluster_id, snapshot, keyspace, column_family, node_id}
• track the overall restore
  Gauge to show:
  • 1 - copying data (transition from any other when SM started to copy the data)
  • 2 - call to refresh sstables on the node (transition from 1 after data is copied)
  • 3 - done (transition from 2 when refresh is done)
    scylla_manager_<clone_restore_or_other_name>_state{source_cluster_id, cluster_id, snapshot, keyspace, column_family, node_id}