Open DaoWind opened 4 years ago
I just found this issue, and want to confirm this issue(after connection loss, it connect agian, but the lock was released). I think it is not better for another MHA manager can start(even it not too much harmful for whole cluster.)
I want to know why use error_count to forbid re-lock monitor when connection loss? I set pass 0 to connect event if connection loss, so it will relock again. Is it OK for this change?