Failed to process passivation queue after DB restart
We just did DB failover test while application is running, there was 2 minutes of outage and Database failover completed. We are able to login to Developer porta but in logs we are getting below errors
java.lang.RuntimeException: Problem during method invocation (invokeEngine)
... 48 more
***Updated by Moderator: Vidyaranjan| Updated Categories***
Pega does not automatically recover for cases like database outage - at the moment (certainly a very desirable enhancement request). You would need to restart the application server to establish clean system. The errors you reported are typical of database outage scenario.
To add to Kevin's comments above. Its highly likely that due to the Database server outage or network issue between your Appserver and Database server, Websphere has raised a StaleConnectionException. Please review the logs (PegaRULES and SystemOut.log) to look for StaleConnectionException prior to the java.lang.NoClassDefFoundError.
When the IBM Websphere Application Server’s Resource Adapter throws a StaleConnectionException, the connection(s) are evicted from the pool. Any application thread requesting a JDBC Connection from IBM Websphere’s JDBC pool when “Pre test” JDBC Connections in the pool feature is enabled (as in your case), Websphere will then try to pretest the JDBC Connection in the pool before handing it out to the requesting application thread. If you have reviwed and confirmed that a StaleConnectionException is being thrown the next question that comes to mind is -
How do I deal with the StaleConnectionException in WebSphere Adapter for JDBC?
Please follow instructions under PDN and IBM Websphere Knowledgecenter articles below to deal with StaleConnections –
Thank you Kevin and Ankur. First of all we are doing DB failover testing, during failover DB was down for 2 minutes.So database was definately not available. But there is no stale connection error in database. We retried the failover without any change and this time there is no such exception in logs. So it looks intermittent. Do you think the connection pool settings need to be tweeked? Below is what my PegaRULES DB connection pool setting looks like.
Thank you for the great responses! Santosh has informed us, moderators, that he has not proceeded with further investigation since the issue has not recoccured. I will be closing this thread to future replies and will reopen the discussion once the investigation can be continued.