I've gotten a second Nagios server setup to work as a failover for a primary server. I think I've been thorough. The secondary server is successfully receiving and processing both passive host and passive service checks. Notifications and both kinds of active checks are turned off. When I stop the Nagios process on the primary server, the secondary fails the freshness check of the primary instance check. However, nothing happens after this, and it seems that I have 2 problems.
1) Even though I have max_check_attempts set to 1 on the master server's "check_nagios" check, it just continues to force active checks when the freshness times out. I expect it to fail hard, and stop checking the freshness. Maybe I'm wrong, though. Maybe the expected behavior here is to get the active host check going, and then the freshness will stop complaining. 2) All of my scripts seem to be lined up. The event handler fires, and I see proper things in the nagios.cmd file. [1246294798] ENABLE_NOTIFICATIONS [1246294798] START_EXECUTING_SVC_CHECKS [1246294820] START_EXECUTING_HOST_CHECKS I know the command file is being processed because I can get the secondary server to force checks from the cgi's. However, none of these things commands ever work, whether I force them from the command line, or from the cgi's. What could be keeping these from taking effect? I've been all over this thing for a couple days now, and I think my eyes are starting to glaze over. The only thing I can think of would be to enable all of these things in the master config file, but then immediately force them "off" when I start up the process. Then maybe it will work to turn them back on later? That can't be right... Desperately, dk ------------------------------------------------------------------------------ _______________________________________________ Nagios-users mailing list Nagios-users@lists.sourceforge.net 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