Restarting the graphing system

[waskinbas]

Probably twice a year the graphing system in my Ubuntu 22.04-based Nagios XI locks up. The server still runs fine, serves a page, does checks and all the other good stuff Nagios is supposed to do, but the graphing system just stops. In the past, I've rebooted the entire server and graphing begins again, leaving a black hole where the outage was. When troubleshooting, I have a (bad?) reputation for putting graphs and charts in my email to prove my findings and help higher-ups understand what did and did not happen, so missing graph information can be pretty annoying. When this happened again earlier this week, I decided to figure out how to write a cron job to restart the graphing system every night a midnight. To start with, I'm targeting rrdcached and npcd, which seem to be closely involved in the graphing system.

Does this happen to anyone else? If so, how do you deal with it? Also, would it be more effective to just reboot the entire server via cron job? I did just upgrade to 2026R1 today, which I'm hopeful might fix that graphing issue as well.

[cdietsch]

Hi waskinbas,

This should not be happening. I would recommend the better solution would be to dig into what is causing the issue in the first place by checking the log files when this happens again. I would also recommend opening a ticket with our support department as they should be able to help you troubleshoot the issue. You can access the Customer Support Portal at https://support.nagios.com/ and open a ticket.

[Sairike]

I’ve run into similar issues occasionally. Scheduling a nightly cron to restart rrdcached and npcd sounds like a reasonable approach — much safer than rebooting the entire server. Hopefully the 2026R1 upgrade helps stabilize it too!