Updated: 22 October 2026 · Applies to: Error
Occasionally, the RPM database on a CloudLinux machine can become corrupted. It can happen when an RPM transaction is interrupted at a critical time. Symptoms of this problem include one of the following programs not responding or freezing:
- up2date
- The CLN alert notification tool (applet in the Gnome or KDE panel)
- rhn_check
- rpm
This problem can also cause a system to stop checking in with CLN. To fix this problem, run the following commands as root:
1. Kill all RPM and yum processes (rhn_check, yum, rpm, rhn-applet):
ps -axwww | grep rpm
In the list of processes, the first number on each line is the PID. For all PIDs listed except for the one associated with grep:
kill -9 <PID>
Repeat the above steps for each of the programs listed.
2. Remove any RPM lock files (/var/lib/rpm/__db*):
rm -rf /var/lib/rpm/__db*
3. Rebuild the RPM database:
rpm --rebuilddb
If the above steps do not work, please contact technical support for more assistance.
A hung package-manager process can also leave a stale lock file behind even after you kill it, so if a fresh yum or rpm command still refuses to run, check for and remove any leftover lock file before trying again.
If killing the process and clearing any lock file still does not help, check whether the server is simply low on free disk space or memory, since yum and rpm can both hang indefinitely while waiting on a resource that never frees up.
