Add validation stage to the 1-1 restore service

[karol-kokoszka]

First step of the 1-1 restore performed by Scylla Manager is to validate that the destination cluster topology matches the source cluster topology.

Scylla Manager must:

• validate that all the nodes have an access to the backup location
• validate that the mapping includes all the nodes
• validate that the source node belongs to the DC and rack defined by the mapping
• validate that the destination node belongs to the DC and rack defined by the mapping
• validate that the destination node has the equal amount of CPUs to the source node amount of CPUs
• validate that the destination node storage size is GE than the source node storage size
• validate that the destination node is the owner of the same tokens like the source node

This process must be parallelized per destination node (or source node).