There NRPE check is a good start for this dependency. There are some checks out there for apache2 health, but our servers aren't configured to serve up the apache "server-status" page which they rely on. Ken Raffenetti Systems Administration Associate MCS Division - Argonne National Laboratory On 05/24/2012 10:15 PM, Schmitz, Corby B. wrote:
and as we all know, a host can be pingable and still not up, which makes it hard to rely on ICMP to do the dependencies which is why I like the NRPE check. If I can't poll the service that should be there, I can assume any other services being down are related.
Corby Schmitz Sent from my mobile office
On May 24, 2012, at 22:10, "Ken Raffenetti" <[email protected]> wrote:
Yeah, I remember not being able to make a service dependent on a host, only another service. At the least, I can monitor Apache on those servers and make all the vhosts dependent on it.
Ken Raffenetti Systems Administration Associate MCS Division - Argonne National Laboratory
On 05/24/2012 10:07 PM, Schmitz, Corby B. wrote:
if I remember correctly, you have to create dependencies for both the hosts and the services that run on them. Nagios segments these, which is sometimes great, and other times very painful.
Corby Schmitz Sent from my mobile office
On May 24, 2012, at 21:59, "Ken Raffenetti" <[email protected]> wrote:
A proper definition of the dependencies for owney/wind/wilbur's vhosts will go a long way, too. I'll make a ticket and get it in my queue.
Ken Raffenetti Systems Administration Associate MCS Division - Argonne National Laboratory
On 05/24/2012 09:49 PM, Max trefonides wrote:
FOOO! FIE! It was a desperate attempt tpo gain some scotchy bounty. I knew it was half baked.
-- Max Trefonides MCS Systems [email protected]
On May 24, 2012, at 9:42 PM, Schmitz, Corby B. wrote:
I hate exchange. My original note was gobbled because perl is not an allowed attachment. Anyway.
Unless you hacked that file up to work with the latest version of nagios,it won't do what you want. In fact, it won't do anything as they changed the format of the state files and the permissions on the command file. In peer:/usr/local/bin/ack.pl you can find the version that does work with nagios 3.2. I have been using this with the CHIMAN install on fluke for about 8 months with great success. It is based on the same file that Max grabbed from the web (thank you Santiago Velasco - sanxiago.com). I added the functionality to ack all (without having to type 'y' over and over, use the fancy A for ACK this crap already). I added the appropriate locations for the files and ran it to ACK remaining alerts to verify that things work correctly in a packaged install, which they do.
Come on max, you need to test your code before you raise your hand :)
-- corby
On 5/24/12 8:40 PM, "Trefonides, Adam A." <[email protected]> wrote:
that's
peer:/sandbox/maxadam/bin/mass_scheduler.pl --
Max Trefonides MCS Systems [email protected]
On May 24, 2012, at 8:39 PM, Max trefonides wrote:
Look in /sandbox/maxadam/bin
script is mass_scheduler.pl
I haven't completely tested it yet.
-- Max Trefonides MCS Systems [email protected]
On May 24, 2012, at 8:16 PM, Craig Stacey wrote:
A bounty of beer, scotch, or other reward to the first person who implements a bulk "acknowledge" on peer.mcs.anl.gov (and any other nagios instance I'm monitoring).
-- Craig