Nagios Support Forum

I noticed that we have about 200 of these rrd does not exist errors. They show up as unknown in the dashboard. the specific error is /var/lib/mrtg/XXX.XXX.XXX.XXX_8.rrd does not exist. (IP removed), we are running the current version of Nagios XI, and have about 2000 monitored devices. I went through lots of forum posts, and found similar issues, but the fixes don't seem to improve my situation. Any help would be great.

When this happens for us, it's been because there are hosts with duplicate names - one uppercase and one lowercase. I opened a support case awhile back and they identified this as the issue. Try looking in /etc/mrtg/conf.d/ on the Nagios server for duplicate host name .cfg files

I ran into something very similar on a larger XI setup and in my case it wasn’t Nagios itself, but missing/corrupted RRD files from MRTG.
az games
If you’re seeing /var/lib/mrtg/...rrd does not exist, usually it means:

the RRD file was never created (polling issue), or
it got deleted/corrupted at some point

A couple things that helped me track it down:

Check if MRTG is actually running and updating:
ps aux | grep mrtg and look at /var/log/mrtg.log (or cron jobs)
Try manually running MRTG for one of those IPs and see if it recreates the .rrd
Verify permissions on /var/lib/mrtg/ (I had a case where files weren’t being written due to ownership issues)
If using SNMP for those devices, confirm they still respond — if polling fails, RRDs may never get generated

In my case, a bunch of devices had SNMP timeouts, so MRTG never created the RRDs, and Nagios just kept reporting them as missing.

Do the affected devices still return valid SNMP data, or are they maybe offline/intermittent?

It seems like this issue is happening every time I add a new device. When I first ran into the error, I threw it into good old ChatGPT, which pointed me to an SNMP v2c reference in the config for that specific device. Since we use SNMPv3, I followed that suggestion and removed the public@ip, and that seems to have fixed it. The problem is I’m using the switch wizard, where there’s very little to go wrong when adding devices, yet this still appears to happen each time.I end up having to update every device manually. With around 5,000 devices, that’s pretty frustrating and not really practical. I’m not entirely sure why the wizard is doing this (or what im doing wrong), but I’m considering opening a ticket to get some help with it.