Seeing Segmentation Faults when Icinga tries to add notification history to the db

Icinga 2
1

Hi All, we’re having a hell of a time trying to figure this one out. Icinga seems to segmentation fault when it tries to update the notification history. The attached image shows the log messages leading up to that point. It never seems to segfault on the same service, but it is always a check where it thought it was “critical” and it has changed to “ok”

segfault
segfault.png3360×658 779 KB
. After icinga restarts itself, it seems to happen every minute or so. However, if we delete the icinga2.state file (which ends up being ~4MB) Icinga seem to stay up around 10 minutes.

We’re running v2.4.4-292
Enabled features: api checker command ido-mysql mainlog notification
Please let me know if I can provide any additional information.

2

never experienced something like this myself, but there are some github issues for the search for “segmentation fault notifications”: https://github.com/Icinga/icinga2/issues?utf8=✓&q=segmentation+fault+notifications

Also, if

really is the Icinga 2 version you are using, it would be best if you update before. I’m pretty sure the problem will be gone after the update to a up-to-date version.
Version 2.10.5 ist the current release.

3

Right, that’s a really old snapshot version which generally remains unsupported. The -292 says that 2.4.4 + 292 Commits on top.