Details
-
Bug
-
Resolution: Fixed
-
Major - P3
-
None
-
None
Description
There is a logic error in BaseCluster#selectServerAsync implementation. If the WaitQueueHandler is about to or already awaiting the next phase with the full heartbeat frequency timeout (default of 10 seconds) and a new request arrive to select a currently unavailable server, the wait queue handler is not notified, and therefore won't wake immediately to try to force a connection to the requested server.