This section describes how to handle migration job failures.
For more information about troubleshooting Zero Downtime Migration and known issues in the current release, see the Zero Downtime Migration Release Notes.
Handling Migration Job Failures
If your migration job fails, the following solutions can help you discover the issue.
If your migration job encounters an error, refer to the migration job output logs, Zero Downtime Migration service logs, and server-specific operational phase logs present at the respective source or target database servers.
If the migration job encounters an exception (that is, fails) then the logs can provide some indication of the nature of the fault. The logs for the migration procedures executed in the source and target environments are stored on the servers in the respective source and target environments. The Zero Downtime Migration command output location is provided to you when the migration job is run with the ZDMCLI
migrate database command. You can also find the log file location in the output of the ZDMCLI
query job -jobid job-id command.
Determine which operational phase the migration job was in at time of failure, and whether the phase belongs to the source or target. Check the Zero Downtime Migration service host log at $ZDM_BASE/crsdata/<zdm_service_node>/rhp/rhpserver.log.0 and access the respective source or target server to check the log associated with the operational phase in /tmp/zdm-unique id/zdm/log.
If the Zero Downtime Migration service does not start, then check the Zero Downtime Migration service logs for process startup errors to determine the cause of the error reported. The Zero Downtime Migration service log can be found at $ZDM_BASE/crsdata/zdm_service_node/rhp/rhpserver.log.0.
If a migration job fails, you can fix the cause of failure and re-run the job. Note that when you re-run a migration job, a new job ID is assigned to the job. Make sure you check the new job and logs for monitoring further progress.