It seems daily the cluster becomes non responsive. The error I get from the php application is "Query failed: Lock wait timeout exceeded; try restarting transaction". My setup is 2 servers running the API and the storage node on each, and then another server running only the management server. There is a Netscaler load balancer in front of the API servers balancing the connections.

When the issue occurs it seem there are more connections then queries. I would say there are probably about 3 times more connections then queries. As far as the traffic goes each server currently does about 100 queries per second, but this will go up as we add more databases to the cluster.

Thanks in advance for your help.