Today, we’re going to address an issue where we cannot edit a backup policy or view the retention policy. Let’s consider an example. We create a retention policy named ‘Non Production Retention’. Once created, we can see this newly formed retention policy. However, due to an incorrect update in the MCS database, we encounter some issues. Upon re-logging into the webpage and navigating to setting, we’re met with a ‘bad response’ error message. The retention policies are not visible, they keep loading, and when we select the backup policy, we get an error stating ‘failed to get policy detail’.
To understand this better, let’s look at the logs over a CLI session. Use the View editor to read the AUI log named ‘mc-rest-api.log’. Press ‘Shift G’ to reach the last line of the log file and use the up arrow key to scroll upwards. Here, we can see the 'No Enum constant' error that caused the issue. Open the MCS Database using the shown command. We see that the retention policy has stored the unit value as zero. This value stands for hours, while the retention policy only has days, weeks, months, and years.
Java does not have a constant value set for the unit of hours, therefore ‘hours’ is not a valid constant. This results in an error saying ‘no enumeration constant’. To resolve this, we first run the MCS flush, using mcserver.sh --flush' then stop the MCS service using 'mcserver.sh --stop --force'. Please note, restarting MCS will not impact any running jobs. However, any waiting clients will be terminated and the log seen in the Activity Monitor will be cleared.
Additionally, the scheduler service will be stopped once the MCS is down. Next, we open the connection to the MCS database 'dbmaint.sh --db=on --dbname=mcdb' Now update the Retention Policy unit to 1 using the 'psql -d mcdb -p 5555 -c' command and the argument "update retention_policies set unit=1 where unit=0;". Verify the change using the same psql command and the argument "select id,name,duration,unit from retention_policies;" Start the MCS service using 'mcserver.sh --start' and verify the status of the MCS service using the command 'dpnctl status mcs' Since the MCS was restarted, the scheduler service will be down.
We start the scheduler service 'dpnctl start sched' and verify its status 'dpnctl status sched'. Now, let’s log back into the webpage and navigate to the Retention tab. We no longer see the error message and the retention policies are now visible. Also, no errors are seen while editing the backup policy. For more information on this topic, please check out the Dell Knowledge Base Article 200096.
Thanks for watching.
