nagios3 crashes with livestatus and downtimed checks

[Brad Marshall]

Issue
-------
When nagios3 is configured to have livestatus from check-mk-livestatus as a broker module, and checks have a downtime applied to them it will crash when the logs rotate. This shows up in /var/log/nagios3/nagios.log as:

   [1445238000] Caught SIGSEGV, shutting down...

Steps to reproduce
-----------------------------
* Install nagios3 and check-mk-livestatus

* Edit /etc/nagios3/nagios.cfg to enable livestatus:

    broker_module=/usr/lib/check_mk/livestatus.o /var/lib/nagios3/livestatus/socket

* To speed up testing, edit /etc/nagios3/nagios.cfg to set:

    log_rotation_method=h

This will cause log rotation to occur hourly, rather than weekly or daily.

* Restart nagios to apply these fixes.

* Apply a downtime on any host or service to last until the top of the next hour

* Wait until that time, and see that nagios crashes with the SIGSEGV error.

Solution
-------------
After some searching around I found a patch at http://lists.mathias-kettner.de/pipermail/checkmk-en/2013-December/011087.html which I have applied to nagios source and thrown up into a PPA at ppa:brad-marshall/nagios for testing. Reproducing the steps above but with the upgraded package means it no longer crashes.

The testing was done with latest patched Trusty, specifically:

$ lsb_release -d
Description: Ubuntu 14.04.3 LTS

$ dpkg-query -W nagios3
nagios3 3.5.1-1ubuntu1

$ dpkg-query -W check-mk-livestatus
check-mk-livestatus 1.2.2p3-1

Please let us know if you need any further information, or any testing done.

[Robie Basak (racb)]

Patching nagios3 is the wrong thing to do and may cause other things to break by changing nagios' ABI. The bug is in check-mk which assumes an incorrect nagios3 ABI. See duplicate bug.

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.