Troubleshooting

As explained previously, the output of pcs status under Failed actions will show you whether there are problems with the cluster resources and provide information as to what you should do in order to fix them.

Here is an example:

• exit-reason='Config /var/lib/mariadb_drbd1/my.cnf doesn't exist': Make sure the configuration file for MariaDB exists and is identical on both nodes.

• exit-reason='Couldn't find device [/dev/drbd1]. Expected /dev/??? to exist': The DRBD device was not created correctly. Review the instructions and try to create it.

As you can see, the exit reason will give you valuable information to troubleshoot and fix the issues you may have. If, after verifying the conditions outlined in the error messages, you are still experiencing issues with a particular resource, it is useful to clean up the operation history of a resource and redetect its current state:

pcs resource cleanup [resource name]

From Kamran, a real world problem scenario, which happens when the...