Nagios Hostgroup lost half of it's devices?

[benhank]

Perplexing issues.
1.I had a hostgroup that contained 514 objects. Suddenly it contained 129 objects. I can verify that nobody made any changes to the system. I do know the time period in which the issue must have happened.
I have done the following:
Gotten a list of the 514 objects, re added them to the hostgroup.cfg via command line and did a

Code: Select all

service nagios restart

the objects then showed up in the hg, but then I did an "apply config", and the numbers dropped again.
So I tried again this time doing a database repair 1st. This time a whole bunch of objects I'd deleted came back in ccm as "unsync'd objects", and the issue persisted.
I then deleted the host group, and repeated the db repair and adding the objects via the command line. This time I only got 124 objects BEFORE apply config.
so I recreated the hostgroup file and tried to import it in ccm\tools \import.
nothing no import.
2. While trying to fix this, I noticed than in the legacy ccm, if I looked at a hosts properties, The host would show up as no being a member of the HG. When I looked at the same host in the new CCM, it was listed as a member of the hostgroup this was the case for about sixty of my hosts.


My concern is that even if I manually readded all 514 hosts, I don't know how the problem happed in the 1st place and if it will happen any other hg's.

[mguthrie]

Gotten a list of the 514 objects, re added them to the hostgroup.cfg via command line and did a : service nagios restart

Note the bold print at the top of the .cfg files that say "Do not manually edit this config files, your changes will be overwritten by NagiosQL"

You have to make the changes in the CCM, and then Apply Configuration, otherwise the hand-edits to the config files will always be overwritten every time you Apply Configuration. The only place that it's safe to edit configuration files by hand is the /usr/local/nagios/etc/static directory.

[benhank]

even if I import the files?
i can take the hit and re add all the hosts, but I am really concerned about what caused this.

[mguthrie]

You probably can import those configs, that should work. I don't recommend doing it that way on a regular basis for config changes, but it should save you from having to redo that work in the CCM.

[benhank]

now, when I do an "apply config" it finishes, but when I go back to ccm/hostgroups i am still prompted to apply the config for changes to take affect.

[scottwilkerson]

If you go to Admin -> System Profile

Are the system time and php time synced and correct?

[benhank]

now that you mention it, no. the system and server time are synced, but both are off by 5 mins.

[scottwilkerson]

do the timezones and times match in Admin -> System Profile

[benhank]

Code: Select all

Server Port: 80
Date/Time
PHP Timezone: America/New_York
PHP Time: Mon, 10 Dec 2012 13:27:15 -0500
System Time: Mon, 10 Dec 2012 13:27:15 -0500

thats what i have

[scottwilkerson]

So after you Apply Configuration from the CCM, it still says that you need to apply configuration for changes to take affect?

Can you post a screenshot?