Hello.

I recently found a bug in the LXC container template for the currently 
available CentOS 7 
LXC container template.

I found this with great surprise after a specific systemd-run service, 
zabbix-agent, was 
configured to use /var/run/zabbix/zabbix_agentd.pid for its PID file, but 
systemd couldn't 
read the PID file from /run/zabbix/zabbix_agentd.pid as it normally would.

The reason:

In CentOS 7 proper, /var/run is a symlink to ../run, effectively making /run 
the actual 
statefile holding directory which is usually tmpfs. Because that symlink didn't 
exist, it was 
storing the statefiles on the actual filesystem and not /run where it belongs.

After I manually replaced /var/run with the appropriate symlink to ../run, 
zabbix-agent 
was started and running just fine with no further issues.

Yes, this was also an error in my configuration file for Zabbix, but it's also 
an issue in the 
fact /var/run is supposed to be a symlink to ../run as well, double-edged sword.

-- 
Eric Renfro

_______________________________________________
pve-user mailing list
[email protected]
http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-user

Reply via email to