On 16-12-2004 at 12:25, Moritz Both wrote:
What I can do is putting all nagios messages into one queue via postmaster filter, but it would be good if all nagios messages concerning one problem (i.e. notifications "warning", "critical", "acknoledgement", "ok") would go into the same ticket.
I would write a Perl script to pipe notifications in existing tickets or in new ones, matching two variables: alert type (PROBLEM or RECOVERY) and service name. New tickets should be created if no tickets are currently open with the same variables. Also, a nice script could modify auto-responses to trigger some kind of automatic outage disclosure (say: "we will get back to you ASAP; in the mean time note that we're experiencing problems on a webserver, check this page to verify the status of your account..."). - Alessandro