<html xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns="http://www.w3.org/TR/REC-html40"> <head> <META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=us-ascii"> <meta name=Generator content="Microsoft Word 11 (filtered medium)"> <style>  </style> </head> <body lang=FR link=blue vlink=purple> <div class=Section1> Hi,<o:p></o:p> <o:p> </o:p> In my service dependency config file, I have this kind of thing :<o:p></o:p> <o:p> </o:p> define servicedependency{<o:p></o:p> hostgroup_name *,!NON_NRPE_NODES<o:p></o:p> service_description nrpe daemon<o:p></o:p> dependent_service_description var partition freespace,tmp partition freespace,root free space,quattor daemons,home partition freespace,firewall<o:p></o:p> execution_failure_criteria u,w,c<o:p></o:p> notification_failure_criteria u,w,c,p<o:p></o:p> }<o:p></o:p> <o:p> </o:p> The hostgroup NON_NRPE_NODES is excluded for the dependencies, because it’s not managed the same way as other nodes.<o:p></o:p> You could probably generate a group called “NRPE NODES” or something like this, and instead of using * like me, define this service dependency on this hostgroup only.<o:p></o:p> <o:p> </o:p> I’m using Nagios 3.0b7 at the moment, I don’t know which version you’re using nor if this was available in Nagios 2.10 though…<o:p></o:p> Hope this helps ?<o:p></o:p> <o:p> </o:p> Regards<o:p></o:p> <o:p> </o:p> -----Original Message----- From: nagios-users-bounces@lists.sourceforge.net [mailto:nagios-users-bounces@lists.sourceforge.net] On Behalf Of Hari Sekhon Sent: Thursday, December 20, 2007 6:17 PM To: Nagios Users mailinglist Subject: [Nagios-users] Remote Execution Service Dependencies on Same Host's Remote Execution Method <o:p> </o:p> Hi,<o:p></o:p> <o:p> </o:p> When there is a problem on a host, and you have lots of services <o:p></o:p> being check through a remote execution method like Nrpe, what do you do <o:p></o:p> if Nrpe breaks or times out or something? You get flooded with alerts <o:p></o:p> for all your systems. So, the solution to this is to create service <o:p></o:p> dependencies that point back to one Nrpe service on the same machine, <o:p></o:p> right? This way you only get 1 alert per machine instead of 10 alerts <o:p></o:p> per machine...<o:p></o:p> <o:p> </o:p> Now the trick is, how do you go about creating a rack of service <o:p></o:p> dependencies for all the remotely checked services via Nrpe for every <o:p></o:p> host to depend on 1 Nrpe check on the same host?<o:p></o:p> Bearing in mind that you don't want to be a mindless drone and create <o:p></o:p> this by hand, and you certainly don't want to maintain this by hand for <o:p></o:p> every host and all of their services...<o:p></o:p> <o:p> </o:p> What do/would you do?<o:p></o:p> <o:p> </o:p> -h<o:p></o:p> <o:p> </o:p> -- <o:p></o:p> Hari Sekhon<o:p></o:p> <o:p> </o:p> <o:p> </o:p> -------------------------------------------------------------------------<o:p></o:p> This SF.net email is sponsored by: Microsoft<o:p></o:p> Defy all challenges. Microsoft(R) Visual Studio 2005.<o:p></o:p> http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/<o:p></o:p> _______________________________________________<o:p></o:p> Nagios-users mailing list<o:p></o:p> Nagios-users@lists.sourceforge.net<o:p></o:p> https://lists.sourceforge.net/lists/listinfo/nagios-users<o:p></o:p> ::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. <o:p></o:p> ::: Messages without supporting info will risk being sent to /dev/null<o:p></o:p> </div> </body> </html>