[feature] add api: stopslavemonitor & startslavemonitor

[TCeason]

A Three-node cluster, if two slave nodes SQL thread exceptions are caused for some network reason, slave nodes will directly stop slave.

This causes a MySQL semi-sync ERROR because all query requests on the master node are waiting.

Master semi-sync variables: rpl_semi_sync_master_timeout=1000000000000000000 rpl_semi_sync_wait_no_slave=ON rpl_semi_sync_master_wait_point=AFTER_SYNC

Maybe xenon needs two new API: stopslavemonitor & startslavemonitor

Detailed description:

These two apis can only be executed when raft role is IDLE.

xenoncli mysql stopslavemonitor

When a slave executes stopslavemonitor, this slave will stop slave status monitor.

xenoncli mysql startslavemonitor

When a slave executes slaveslavemonitor, this slave will start a slave monitor.