Hi All,
We have been experiencing some strange behaviour in one of our three environments. From the ODI client, we can see the output messages when it connects to the master repository and work repository. For the connection to Master Repository, the output is usually in the following format:
INFO: New data source: [ODI_REPO/*******@jdbc:oracle:thin:@host:PORT/ServiceName]
Jul 04, 2017 10:51:31 AM oracle.ucp.common.UniversalConnectionPoolBase
INFO: inactive connection timeout timer scheduled
However, the connection to Work Repository usually translates the connection string to a long connection descriptor like this:
INFO: New data source: [ODI_REPO/*******@jdbc:oracle:thin:@(description=(address=(host=HOST)(protocol=tcp)(port=1521))(connect_data=(service_name=SERVICE_NAME)(server=dedicated)))]
Jul 04, 2017 9:35:35 AM oracle.ucp.common.UniversalConnectionPoolBase
INFO: inactive connection timeout timer scheduled
This is the current behaviour of DEV and PROD. But it is currently different on TEST where the output of the connection to Work Repository is exactly the same string as the connection to the Master Repository (is not translated into a long descriptor).
Do you have any idea about why this is happening? It should be related to a different database setup somewhere.
The only issue I am facing is that the connection is closed/lost after few hours of inactivity. Everyday in the morning the connection is lost from ODI and I have to disconnect and connect again (sometimes I have to close the client as well because it becomes not responsive).
From the other environments. I can keep the connection alive for days or even weeks without any issue.
Any ideas?
Thanks!