Nagios Support Forum

Hi,

Running Nagios XI 5.2.9 on Linux RHEL 6 x86-64.

When I restart Performance Grapher (NPCD) from the GUI (XI System Component Status), it comes back in error status (red exclamation mark).

I did some troubleshooting and found that the ownership of the PID file seems to be the problem.

1. after a reboot all systems are 'green'. NPCD runs with the following process info and the PID file is owned by nagios.nagcmd:

$ ps aux | grep npcd
nagios 2245 0.0 0.0 368892 1028 ? S 10:28 0:00 /usr/local/nagios/bin/npcd -d -f /usr/local/nagios/etc/pnp/npcd.cfg

$ ps -eo euser,ruser,suser,fuser,f,comm,label|grep npcd
nagios nagios nagios nagios 5 npcd system_u:system_r:initrc_t:s0

$ ls -l /usr/local/nagiosxi/var/subsys/npcd.pid
-rw-r-----. 1 nagios nagcmd 4 Nov 22 10:28 /usr/local/nagiosxi/var/subsys/npcd.pid

2. After restart of NPCD in the GUI:

$ ps aux | grep npcd
nagios 13876 0.0 0.0 162040 948 ? S 11:02 0:00 /usr/local/nagios/bin/npcd -d -f /usr/local/nagios/etc/pnp/npcd.cfg

$ ps -e -o euser,ruser,suser,fuser,f,comm,label|grep npcd
nagios nagios nagios nagios 5 npcd system_u:system_r:initrc_t:s0

$ ls -l /usr/local/nagiosxi/var/subsys/npcd.pid
-rw-r-----. 1 root root 5 Nov 22 11:02 /usr/local/nagiosxi/var/subsys/npcd.pid

3. Changing ownership fixes the error: chown nagios. /usr/local/nagiosxi/var/subsys/npcd.pid


The privilege options in /usr/local/nagios/etc/pnp/npcd.cfg are:

# Privilege Options

user = nagios
group = nagios

NPCD is restarted as follows:

# tail -f /usr/local/nagiosxi/var/cmdsubsys.log
...................................................PROCESSING COMMAND ID 209...
PROCESS COMMAND: CMD=113, DATA=a:1:{i:0;s:0:"";}
CMDLINE=sudo /usr/local/nagiosxi/scripts/manage_services.sh restart npcd
NPCD Stopped.
NPCD started.
OUTPUT=NPCD started.
RETURNCODE=0

So it appears that /etc/init.d/npcd, which is run at boot, sets the PID file correctly. But /usr/local/nagiosxi/scripts/manage_services.sh restart npcd does not.

Is this something that can be tweaked?

How was your install done? Has it been modified in any way?

My clean install of XI 5.2.9 on CentOS 6 gives the following: I am unable to account for the behavior you're seeing.

/usr/local/nagiosxi/scripts/manage_services.sh uses the same init script.

That snapshot helped.

Changing the permissions on the PID file from 640 to 644 made the error go away, so it appears to be a umask setting.

You can go ahead and close the thread.

Good to hear your issue is resolved. If you have further issues please open a new thread or ticket.