CA PAM Node Disk Utilisation is more than 80%

book

Article ID: 211345

calendar_today

Updated On:

Products

CA Privileged Access Manager (PAM)

Issue/Introduction

Sometimes it is observed that a master node may have high disk usage, over 80% with no apparent reason. Looking at possible causes, we realized that a database backup was taken daily.  

Possible reasons of a daily backup

Configuration --> Database --> Scheduled backups

This can be done automatically via schedule database backup, that was empty, so was not the reason

or because a patch install, and was not possible either this cause because was not installed a patch every day to the appliance.

 

 

Cause

 

Confirmed that the configuration in Configuration --> Database --> Scheduled backups shows that no scheduled backup.

So we go to the appliance via ssh to check if cron show any job schedule, sometimes the problem is that even if you do not in GUI, it is running and active in appliance.

Environment

CA Privileged Access Manager, can happen in most versions.

Resolution

To correct the problem it is necessary to gain ssh access to the appliance. Please engage Broadcom support to assist you with this.

Once connected through ssh to the PAM node the following actions will help correct the disk space problem

  • Clear old db backups in the GUI (or if for some reason not available, in the default database backup location /var/uag/db/backup). This will help recover disk space
  • In the ssh terminal run crontab -e. The result is similar to the one below

In this particular example the second line shows a daily backup scheduled at 13:30 which should not be there according to the information in the GUI

To correct this situation, do the following:

  • Access the PAM GUI for this node and there go to Configuration --> Database --> Scheduled backups and create a backup like the one shown in the the result of crontab -e
  • Next delete this scheduled backup using the PAM GUI as before
  • Run crontab -e in the ssh terminal to check that the scheduled job has really been deleted

This remediation will both allow the disk space to be recovered and it will also prevent a future occurrence of the same problem.