Wait for the whole replication length parameter when reading replication messages over TCP. Should fix #342
Also fix issue where dbMap was being used instead of the copy in Sync/Dump DBs commands - this would cause issues if a sync was taking place while a statsDB was being added or deleted. Unlikely, but possible.
Wait for the whole replication length parameter when reading replication messages over TCP. Should fix #342
Also fix issue where dbMap was being used instead of the copy in Sync/Dump DBs commands - this would cause issues if a sync was taking place while a statsDB was being added or deleted. Unlikely, but possible.