Storing Transaction Logs In The Database Cause Graceful Server Shutdown Issues With MDB (Doc ID 1544574.1)

Last updated on JUNE 13, 2024

Applies to:

Symptoms

When shutting down a WebLogic Server Cluster, is possible to get core dumps for the related JVMs. The logs complain that a JDBC resource can't be found, although it exists on the console and in the config directory:

The problem occurs when using a Oracle RAC as backend and multidatasources are configured to point to the Database Back End.

Steps to reproduce are listed below:

1. Create a cluster with 2 managed servers.
2. Create a datasource and use this data source for tlog.
3. Create a JMS server on each managed server with JDBC persistence.
4. Create a distributed topic.
5. Deploy an MDB to cluster with the distributed topic target.
6. Send a sample message to the topic and see if the MDB is executing.
7. Gracefully shut down the server - it will not shut down because the MDB is still active.
8. Try force shutdown. It will hang for a while, go to critical state, and die.
9. If you have node manager the server will restart automatically.

Changes

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.