On 2011-07-04 15:48, Diana Scannicchio wrote: > Hello, > I would like to have more detailed information about how to install and setup > mod_gearman with Icinga. > Looking into Icinga web pages and instructions I found two different pages > and I am confused:
i'd rather take the official docs provided with mod_gearman itsself. the newly written howtos only make the configs looklike more for Icinga than Nagios. > - > https://wiki.icinga.org/display/howtos/Icinga+with+mod_gearman+on+RHEL+and+Debian this one sources from one of our internal productive setups, thus it was not added to the existing one, as the use case is slightly different. > - > https://wiki.icinga.org/display/howtos/Setting+up+mod+gearman+with+Icinga#SettingupmodgearmanwithIcinga-Install this one is a bit outdated, sorry for the confusion. > I perfectly know and understand that is difficult to keep documentation up to > date, but finding slightly difference instructions for the same actions is > quite confusing (some also referring to nagios). And of course also gearman > site (http://gearman.org ) provides slightly different instructions. Well, these howtos are part of the past months full of setups and different use cases we had at work. so the plan is to finish that (when there's documentation time again!) and also to add further instructions. point is, i can't share all the information in the first place as many of such things may contain internal stuff. but ever since this is a community wiki, feel free to register an accound on www.icinga.org and use that for proper editing and adding information to the howtos yourself. > Which one I am supposed to follow to well cope with Icinga? the first one works well, the second one is jsut a quickinstall without any further notice on the configuration. even more, the first one targets the icinga to be installed on rhel, providing a mod_gearman spec file for packagers, and adding information for a client worker on debian. so you might suspect that this guide isn't yet completed (e.g. adding more workers on various platforms, with different sources and different packages). > I have installed Icinga 1.4.0 on Linux. I have an Icinga instance running on > a central server and another one running on a second server (i.e. distributed > monitoring) that is monitoring tens of nodes and sending to the main one the > results of the check (NRPE/NSCA). This can be considered a testbed. that's the old fashioned way doing it with the provided addons, but not depending on any external application, such as a message queue daemon like gearman would add. > My final system will be composed by ~4000 nodes (all Linux, current kernel is > 2.6.18-238.1.1.el5). > I have to evaluate Icinga and all possible ways to monitor our system, > probably with a central and few distributed servers, to understand if it > copes with our needs and how to migrate to Icinga. the real question would be - how are you planning to implement more than one node, and how is the network being organized. if you are capable of one Icinga master, running the gearmand too, you'll need tcp+udp/4730 to be opened up for the mod_gearman worker clients to allow to connect over there. so if you have various sites with client workers they will all get their data (the check commands and event handlers) asynchronous from the gearmand divided by host/servicegroups, and feed the checkresults back to the gearmand whereas mod_gearman takes care of feeding that to the core finally. such a setup does not require any more Icinga instance than the master itsself, but adds the gearman dependency on that setup. this also works flawlessly with IDOUtils and a centralized Icinga Web than (and Classic UI too). If you are preferring the way of distribution with single Icinga instances on those sites, writing their data into a combined database using IDOUtils, and having Icinga Web extended somewhere, showing such information divided by instance_name, combined into one web ui, that would be another way to go. And of course the way with the nsca master/slave architecture. And somehow Merlin, but I've never tested that voodoo code in test environments for diostributed monitoring. kind regards, Michael > Thank you very in advance for any information you can provide and best > regards, > > Diana > > > > > > > ------------------------------------------------------------------------------ > All of the data generated in your IT infrastructure is seriously valuable. > Why? It contains a definitive record of application performance, security > threats, fraudulent activity, and more. Splunk takes this data and makes > sense of it. IT sense. And common sense. > http://p.sf.net/sfu/splunk-d2d-c2 > _______________________________________________ > icinga-users mailing list > icinga-users@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/icinga-users > -- DI (FH) Michael Friedrich Vienna University Computer Center Universitaetsstrasse 7 A-1010 Vienna, Austria email: michael.friedr...@univie.ac.at phone: +43 1 4277 14359 mobile: +43 664 60277 14359 fax: +43 1 4277 14338 web: http://www.univie.ac.at/zid http://www.aco.net Icinga Core& IDOUtils Developer http://www.icinga.org ------------------------------------------------------------------------------ All of the data generated in your IT infrastructure is seriously valuable. Why? It contains a definitive record of application performance, security threats, fraudulent activity, and more. Splunk takes this data and makes sense of it. IT sense. And common sense. http://p.sf.net/sfu/splunk-d2d-c2 _______________________________________________ icinga-users mailing list icinga-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/icinga-users
