Issues with Nagios Fusion 2012R1.2

[Mitchell]

I am having 2 issues with Fusion currently.

1. The tactical summary does not reflect the latest managed server configuration.
for example, I have 2 XI servers in managed servers configuration (after few deletes and add) but tactical summary is showing 5. (please see attached screenshot)

2. The dashlet is not able to pull latest states from XI server. please see attached screenshot.
The XI server is added to managed servers in fusion. Configured the credentials in configure server. auth_check looks OK. still no data availabe on any of the status dashlets in Fusion.

Regards
Ashish

[mguthrie]

There weren't any new additions to the subsystem poller for 1.2 that I can think of that would have affected that. However, can you run:

Code: Select all

tail -f /usr/local/nagiosfusion/var/poller.log

and see if that reveals anything telling about where the problem might be.

[Mitchell]

Thanks Mike.

looking at logs I was able to fix the issue #2. I missed to grant 'can see all host and services' to the user in XI.

however I still see the issue#1 with 'Tactical Summary' dashlet. I see the additional servers in dashlet, which are no longer in manager server. Where is this dashlet polling the data from?

[root@pnagios01lxv var]# tail -f /usr/local/nagiosfusion/var/poller.log
Saving to DB SID: dlnium
Saving to DB SID: gehgba
2 callbacks run
.........***GET DATA!***
RUNTIME: 93 INTERVAL: 93 POLL INTERVAL: 30
Saving to DB SID: dlnium
Saving to DB SID: gehgba
Polling Recent Alerts...
NAGIOSXI ALERT FETCH pnagios02lxv
COUNT :6
NAGIOSXI ALERT FETCH pnagios03lxv
COUNT :0
2 callbacks run
...***GET DATA!***
RUNTIME: 126 INTERVAL: 30 POLL INTERVAL: 30
Saving to DB SID: dlnium
Saving to DB SID: gehgba
Polling Recent Alerts...
NAGIOSXI ALERT FETCH pnagios02lxv
COUNT :6
NAGIOSXI ALERT FETCH pnagios03lxv
COUNT :0
2 callbacks run
...***GET DATA!***
RUNTIME: 159 INTERVAL: 30 POLL INTERVAL: 30
Saving to DB SID: dlnium
Saving to DB SID: gehgba
Polling Recent Alerts...
NAGIOSXI ALERT FETCH pnagios02lxv
COUNT :6
NAGIOSXI ALERT FETCH pnagios03lxv
COUNT :0
2 callbacks run
...***GET DATA!***
RUNTIME: 192 INTERVAL: 30 POLL INTERVAL: 30
Saving to DB SID: dlnium
Saving to DB SID: gehgba
Polling Recent Alerts...
NAGIOSXI ALERT FETCH pnagios02lxv
COUNT :6
NAGIOSXI ALERT FETCH pnagios03lxv
COUNT :0
2 callbacks run

[mguthrie]

The server deletion is *supposed* to purge the cache for the tactical data, but I'll do some digging and see if there's a bug in there somewhere.

[mguthrie]

Yeah, just confirmed it on my test box. I'll post a temporary hotfix in a bit and fix this in SVN for the next release.

[mguthrie]

Hot fix:

Run the following:

Code: Select all

psql nagiosfusion nagiosfusion
DELETE FROM fusion_tac_data WHERE server_name='theName';
\q

The updated script is attached, replace the following file with the attached.

/usr/local/nagiosfusion/html/includes/utils-servers.inc.php


Fixed in SVN for the next release. Thanks for the heads up!

[Mitchell]

Thanks Mike.

I removed the unwanted server using the sql statement and also replaced the 'utils-servers.inc.php'.

I wanted to test it out. I added a dummy server and removed it to test if the dashlet reflects the changes. It did not work and I had to use the sql statement to remove it. Is that the expected step for now?

Thanks
AShish

[mguthrie]

Just to eliminate as a possibility, do you reload the pages after deleting the server?

Can you verify the following lines exists around 418 in /usr/local/nagiosfusion/html/includes/utils-servers.inc.php

Code: Select all

	//remove from tac_data
	$query = "DELETE FROM fusion_tac_data WHERE server_sid='{$sid}'"; 
	exec_sql_query(DB_NAGIOSFUSION,$query,true); 

I double checked the patch on my local system and I was able to cleanly remove the servers ok. I'm not sure what that wouldn't have worked...

[Mitchell]

Yes. I reloaded the page. also did ^F5 to clear cache just in case.
The file has the code as well.

Do I need to restart anything? I restarted Apache but did not help

[mguthrie]

Ok, I might be losing my mind, because I just tried it again today and the deletion didn't work. (grumble). I'll post the hotfix as soon as it's ready.