Nagios Won't Start - No PID

[laxie16]

Hello,

Recently, our Nagios server running CentOS 5.6 shutoff unexpectedly. Since then, I've been unable to get Nagios to start. Issuing 'service nagios start' will say that Nagios has started, however when I do a 'ps -A' to look at all process ID's, there are no nagios processes. I can view the web monitoring page, but I get an error when trying to view hosts.

[root@nagios var]# service nagios start
Starting nagios: done.
[root@nagios var]# service nagios status
nagios is not running
[root@nagios var]#

I tried upgrading to the latest version, but that did not resolve the issue. I don't see any errors in the nagios.log file, here is what I get when I restart Nagios. It's interesting that it shows the PID, but that process doesn't actually exist.

Anyone have any ideas? Thanks!

Code: Select all

[1349990543] Nagios 3.4.1 starting... (PID=9785)
[1349990543] Local time is Thu Oct 11 17:22:23 EDT 2012
[1349990543] LOG VERSION: 2.0
[1349990543] livestatus: Livestatus 1.1.12p6 by Mathias Kettner. Socket: '/usr/local/nagios/var/rw/live'
[1349990543] livestatus: Please visit us at http://mathias-kettner.de/
[1349990543] livestatus: Hint: please try out OMD - the Open Monitoring Distribution
[1349990543] livestatus: Please visit OMD at http://omdistro.org
[1349990543] livestatus: Removed old left over socket file /usr/local/nagios/var/rw/live
[1349990543] livestatus: Ignoring duplicate logfile /usr/local/nagios/var/archives/nagios-04-27-2012-00.log

[mguthrie]

I'd start by temporarily disabling the livestatus event broker. It'll make it easier to troubleshoot the core engine. Check permissions on everything in your /usr/local/nagios/var directory.

[laxie16]

I'd start by temporarily disabling the livestatus event broker. It'll make it easier to troubleshoot the core engine. Check permissions on everything in your /usr/local/nagios/var directory.

Where can I disable the livestatus event broker? I gave 777 permissions to the /usr/local/nagios/var directory to test, but still saw the same results.

Thanks.

[mguthrie]

There should be a line in your main nagios.cfg file that points to the livestatus broker object. It will look something like:

broker_module=/path/to/livestatus.o /path/to/livestatus.cfg

Then restart nagios

[laxie16]

Disabled livestatus and restarted Nagios. Wasn't able to generate a PID though. I did enable -1 debugging and this is what I get when starting Nagios.

Code: Select all

[1350060106.008403] [001.0] [pid=30148] xrddefault_read_state_information() start
[1350060244.195147] [001.0] [pid=2902] drop_privileges() start
[1350060244.211493] [004.0] [pid=2902] Original UID/GID: 0/0
[1350060244.212238] [004.0] [pid=2902] New UID/GID: 500/500
[1350060244.220239] [064.1] [pid=2902] Making callbacks (type 9)...
[1350060244.220474] [064.1] [pid=2902] Making callbacks (type 9)...
[1350060244.220557] [064.1] [pid=2902] Making callbacks (type 9)...
[1350060244.220570] [064.1] [pid=2902] Making callbacks (type 7)...
[1350060244.245445] [064.1] [pid=2902] Making callbacks (type 7)...
[1350060244.246176] [064.1] [pid=2903] Making callbacks (type 7)...
[1350060244.247532] [064.1] [pid=2903] Making callbacks (type 9)...
[1350060244.301765] [064.1] [pid=2903] Making callbacks (type 26)...
[1350060244.301835] [001.0] [pid=2903] xrddefault_read_state_information() start

[mguthrie]

Do you have anything in /var/log/messages about a segfault error?