Nagios 1.2 not logging?

Crymson · Posted: Thu Nov 04, 2004 2:29 am Post subject: Nagios 1.2 not logging?

I did the Nagios 1.2 ebuild, and it works mostly. I followed the instructions to set up cgi authorization, and set all the paths to the config files, etc, and it starts ok, but nothing is being added to the status.log or status.sav files.

The nagios.log file is filling up with lots of crap, mostly that it can't send mail (because sendmail isn't installed).

When I open the web Interface, I log in, and I get the following on Service Detail (etc)

There doesn't appear to be any service status information in the status log...

Make sure that Nagios is running and that you have specified the location of you status log correctly in the configuration files.

In the upper left, it says:
Logged in as admin
Warning: Monitoring process may not be running!
Click here for more info.
- Notifications are disabled
- Service checks are disabled

So I click "HERE"
Program Start Time: 12-31-1969 17:00:00
Total Running Time: 12726d 1h 26m 13s
Last External Command Check: N/A
Last Log File Rotation: N/A
Nagios PID 0
Notifications Enabled?
NO
Service Checks Being Executed?
NO
Passive Service Checks Being Accepted?
NO
Event Handlers Enabled? No
Obsessing Over Services? No
Flap Detection Enabled? No
Performance Data Being Processed? No

On the bottom, it says:

Process Status Information
Process Status:
WARNING
Check Command Output: Nagios problem: located 2 processes, status log updated 1099531573 seconds ago

Any ideas on WTF is going on? I ran nagios -v /etc/nagios.cfg and it gave me 4 warnings about some hosts not belonging to a group, which I ignored. For the record, I modified some config files that came off a Debian system running nagios 1.2 which WORKED (except that it never found the nagios running process, so we moved it to gentoo, because it had an ebuild) yet it still doesn't work. This time, it can find the process, but no status is being logged, nor do the hosts show up in the web interface.

Please say which config files I need to post, and I will, gladly.

Thanks.
_________________
Knowledge is Power // Power Corrupts // Study Hard // Be Evil

rex123 · Apprentice Joined: 21 Apr 2004 Posts: 272

Not sure, but maybe Debian keeps its log files in a different place to Gentoo. And maybe you have permissions problems.

Do something like this:

Crymson · Posted: Thu Nov 04, 2004 6:44 pm Post subject:

I know the permissions are set correctly, I've gone over that a zillion times now. What's interesting, is that /var/nagios contains *some* log files, but not all. I thought maybe it was logging there, instead of /var/log/nagios, but when I touch the status.log in /var/nagios, and restart the daemon, the file is removed... go fig.

When I get back to the office, I'll take a look at grep output, and the init script.

As for the date, I think the reason it's giving me that, is because there's nothing IN the log file- at least, nothing in the format it's expecting. As I said, there is stuff being added to the nagios.log file (in /var/log/nagios I might add) and if I run check_nagios on that file, I get

Crymson · Posted: Thu Nov 04, 2004 9:53 pm Post subject:

Ok, I fixed the init script to point to the right log directory. That's all done and good.

Yet, Nagios is STILL not creating the status.log file. I think it's the lack of that file that is causing all my problems. Nagios is supposed to be sending stuff to this status.log file, but I don't know WHICH config file governs that.

nagios.log contains stuff about being unable to send messages to the admin, (because sendmail isn't installed). Is there a way to turn host checking back on? It should be able to ping and such, I've run the script from the command line, and it works just fine.

Ideas?? Thanks!!
_________________
Knowledge is Power // Power Corrupts // Study Hard // Be Evil

rex123 · Apprentice Joined: 21 Apr 2004 Posts: 272

Sorry, out of ideas. If the user nagios has permission to create files in /var/log/nagios, which is where your status.log should be, then you should be OK.
The init script removes status.log

Crymson · Posted: Fri Nov 05, 2004 9:23 pm Post subject:

I saw that it deletes the log file for whatever strange reason.

Anyway, as it turns out, the permissions weren't set on /var/log/nagios - which was the root of all my problems. Damn annoying if you ask me, since I was busy checking permissions on every OTHER directory and file.

*sigh*

Thanks again all for your help in narrowing down my stupid and ignorant mistake. The other suggestions were very helpful!!
_________________
Knowledge is Power // Power Corrupts // Study Hard // Be Evil

ewtrowbr · Posted: Fri Nov 05, 2004 9:59 pm Post subject:

when you get the error message "Warning: Monitoring process may not be running! " in nagios, the application is definitely not running.

We have a gentoo nagios implementation monitoring 250 services across 60 servers and it is running great. It can and will work.

erich

Crymson · Posted: Fri Nov 05, 2004 10:09 pm Post subject:

You misunderstand - on the old Debian machine, things WERE working just fine. In fact, I can start the service, and it comes right up, but the check_nagios plugin isn't returning the proper value for some reason. In fact, a ps ax shows nagios as a running process.

It was this problem that caused my boss to decide to use a "clean" install of nagios, and told me I could put it on whatever platform I wanted. So I went Gentoo. Anyway, it's mostly up, just a few minor things to work out now.
_________________
Knowledge is Power // Power Corrupts // Study Hard // Be Evil

rex123 · Apprentice Joined: 21 Apr 2004 Posts: 272