Hi All,
We faced a problem with JDBC adapters in Cluster environment.Adapter services failed in one node of the cluster due to error "[ART.118.5053] Adapter Runtime (Connection): Unable to get a connection to resource ".
All these times, the other node in the cluster did not show any issues.
Below is the connection parameters(same in both nodes):
Minimum Pool Size 0
Maximum Pool Size 100
Pool Increment Size 10
Block Timeout (msec) 60000
Expire Timeout (msec) 60000
Checked DB performance thinking of delay in DB processing,but it was good. Noticed that there are inactive sessions to the DB at this time and so refreshed the connection in this node.Inactive sessions got cleared and for some time(around 6 hrs) started procesing well but same failures started after that. Refreshed the connection again and started processing good. After nearly 12 hours failure started and again refreshed. After this the failures did not occured yet and is working fine for several days.
The volume of the transactions was not changed at this times and all of a sudden issue started.This problem occured 2 times in one month time.
Any of you faced this kind of problems? Any idea about the cause of this strange behavior?
Thanks in advance
ANS
Does firewall exist between IS host and DB? Are you able to ping DB host when problem occurs?
No firewall exists between the systems and was able to ping DB at this time.
Check with your DBA to see if they are okay with a timeout of 60 seconds. Short session expirations can create session thrashing, though I’m not sure if that’s what’s going on here. The expiration should probably be just a little bit shorter than what the DB will use for purging idle connections. E.g. if the DB will close the connection after 2 hours, configure your pool expiration setting to 1:55.