[Tablets] Skip any decommission/remove/terminate node nemesis from running when the number of nodes is not larger than the number of replicas

[ShlomiBalalis]

Issue description

In such nemeses, when trying to drain a node and the total number of nodes is not larger than the number of replicas, the drain process fails:

2024-02-05 11:32:23.970 <2024-02-05 11:32:23.869>: (DatabaseLogEvent Severity.ERROR) period_type=one-time event_id=147b21f0-4531-4dd4-9163-8e006a48fd33: type=RUNTIME_ERROR regex=std::runtime_error line_number=3299 node=longevity-5gb-1h-TerminateAndRemove-db-node-659547b4-1
2024-02-05T11:32:23.869+00:00 longevity-5gb-1h-TerminateAndRemove-db-node-659547b4-1      !ERR | scylla[6009]:  [shard 0:strm] raft_topology - tablets draining failed with std::runtime_error (Unable to find new replica for tablet 645da7f0-c419-11ee-8ee9-26c98a3da6ca:1 on daf03d5e-691b-448c-9a59-4dc058c920fd:1 when draining {daf03d5e-691b-448c-9a59-4dc058c920fd}). Aborting the topology operation

2024-02-05 11:32:23.976 <2024-02-05 11:32:23.869>: (DatabaseLogEvent Severity.ERROR) period_type=one-time event_id=147b21f0-4531-4dd4-9163-8e006a48fd33: type=RUNTIME_ERROR regex=std::runtime_error line_number=3307 node=longevity-5gb-1h-TerminateAndRemove-db-node-659547b4-1
2024-02-05T11:32:23.869+00:00 longevity-5gb-1h-TerminateAndRemove-db-node-659547b4-1      !ERR | scylla[6009]:  [shard 0:strm] raft_topology - Removenode failed. See earlier errors (Rolled back: Failed to drain tablets: std::runtime_error (Unable to find new replica for tablet 645da7f0-c419-11ee-8ee9-26c98a3da6ca:1 on daf03d5e-691b-448c-9a59-4dc058c920fd:1 when draining {daf03d5e-691b-448c-9a59-4dc058c920fd}))

We should add a check at the start of those nemeses, and skip them if needed

[roydahan]

@ShlomiBalalis This is maybe an enhancement but it shouldn't block you in any case.
If it's something that "blocks" you, the first thing you should do is to change the case or to add a new case to have the configuration you want.

[roydahan]

@ShlomiBalalis I don't know what is the suggested procedure from product, they should supply you the procedure based on use cases.
(E.g. replacing live node, dead node, shrink cluster).

When ALTER will be supported this case should use ALTER, this will also cover you the case for ALTER keysapce with RF change.

[bhalevy]

@ShlomiBalalis please close if everything was dealt with

[yarongilor]

Should be addressed in #7625

[swasik]

Should be addressed in #7625

@yarongilor #7625 is merged now so could we close this one?