On 08/05/2010 04:16 PM, Linus Gasser wrote: > Le 05.08.10 18:38, Herb J. a écrit : > >> That would be considered a standard distributed Nagios setup. We do that >> in our Nagios cluster. >> >> We have a number of remote Nagios collectors in different locations. >> They all forward their service check results back to a central >> monitoring server. That server doesn't actually run any host or service >> checks (other than to monitor the remote collectors), but that depends >> on how you configure your Nagios installation. Each remote collector is >> fully independent in the fact that they have the standard Nagios web >> interface, they schedule their own service checks, and they can generate >> notifications for failed hosts/services (we decided against that and >> only have our central collector generating notifications). The data is >> passed back using OCP_daemon to improve performance, which feeds data >> back to the central collector using NSCA. >> > OK, I'll try that again, then. I've been looking at NagiosCenter View, > but I don't manage to make the configuration as it should be. So up to > NSCA, then ;) > > >> However, due to the size of our setup, without heavy use of templating >> and fully automated configuration file management, it would be a >> nightmare to maintain. >> > Hmm, in fact, do you have to configure the central collector to let him > know what the other collectors are doing? >
All collectors are configured with the same check commands and templates, but the remote collectors only know about their specific list of hosts and services to monitor. The central collector knows about all hosts, services, service check commands, etc. (Since Nagios ignores passive data for hosts and services it doesn't recognize, it has to know about everything being monitored.) However, the central collector is configured to not run any active checks (except for monitoring the other collectors) and is configured to accept passive check data. The remote collectors are configured to run active service checks on their list of servers (ignoring passive data since they don't receive any) and does not generate any alerts or notifications. All of this is done by having the central collector use slightly modified host and service templates. > Thanks, > > Linus > > ------------------------------------------------------------------------------ > This SF.net email is sponsored by > > Make an app they can't live without > Enter the BlackBerry Developer Challenge > http://p.sf.net/sfu/RIM-dev2dev > _______________________________________________ > Nagios-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/nagios-users > ::: Please include Nagios version, plugin version (-v) and OS when reporting > any issue. > ::: Messages without supporting info will risk being sent to /dev/null > ------------------------------------------------------------------------------ This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev _______________________________________________ Nagios-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/nagios-users ::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. ::: Messages without supporting info will risk being sent to /dev/null
