2

at our company we have installed a Icinga together with PNP4Nagios. We've got a rare problem with the perfdata files PNP4Nagios creates.

Each couple of weeks, we notice that our icinga server stops working correctly because it runs out of inodes. I've been looking around, and I've noticed that the problem is here:

/usr/local/pnp4nagios/var/spool

Each couple of weeks, this folder contains over 300k perfdata files of this type: perfdata.1336058870

I think it must be a PNP4Nagios problem, because I think that normally each "X" time, PNP4Nagios should override existing perfdata files to NOT generate new ones.

Do you have any idea on what I could change or what I must do to solve this problem?

Thanks in advance.

quanta
  • 50,327
  • 19
  • 152
  • 213
Kevin Maschke
  • 97
  • 1
  • 7

1 Answers1

4

The way that nagios/icinga and pnp4nagios work is that icinga puts the perfdata file into a directory (in your case, /usr/local/pnp4nagios/var/spool). Then pnp4nagios reads the file, processes the contents, and deletes it. The second part seems to be the problem.

First, is pnp4nagios working? If so, it can read the perfdata files. Check the permissions on the spool directory, and make sure that whatever user you're running pnp4nagios (or npcd) as (by default nagios, even on an icinga setup) has write permissions to that directory.

You may also find some guidance from the contents of /var/log/pnp4nagios - it'll log there what it tries to do and when it fails.

NathanG
  • 1,326
  • 10
  • 14