Overview

In some cases, it's possible that on a device that has been running for a long time, that the files in /var/opt/tms/stats become large and take up a lot of space in the /var partition. Since /var is limited to 10GB, it can be a problem if 6-7GB is used up by /stats, leaving very little room for dead.letters, sysdumps, snapshots, etc. This issue will also cause a number of graphs on the web UI to become empty.

This article provides the recommended way to clear this data to make more space for when it is needed.

Process

For reference, the files in /var/opt/tms/stats are CHD and sample data files, which store the monitoring data for things like Monitor > Optimization, Monitor > Interfaces, Dashboard > Link Utilization, etc. (an old graph style which is generally red and blue colored).

In version 6, the size limits for the files in /stats were set extremely large but were later set to realistic sizes in version 7. It appears that if the device was initially running version 6, then was upgraded to version 7, that the limits will remain at the considerable size they were set to in version 6.

The way to enforce the limits from version 7 is to recreate the files. This would mean that the monitoring data for the graphs mentioned above will be lost.

To delete the files

1. Access the Exinda shell.
   

   Important: if you're prompted for a License key while running _shell command, you need to Activate the Restricted Commands License, open a Support Ticket and ask them to activate it.

2. Terminate statsd and go to shell.
   

     pm process statsd terminate

3. Navigate to the directory and delete all CHD and sample data files, then go back to CLI:

     cd /var/opt/tms/stats
     rm -rf sample-*
     rm -rf chd-*
     cli

4. Now, restart statsd and go back to shell to confirm that all the files have been recreated.
   

     en
     conf t
     pm process statsd restart
     _shell