Escalations between two states are not working


#1

[size=150]I posted this in the wrong forum so I’m posting it here instead.[/size]

I am having an issue that I can’t find an answer to in these forums. I have a service that checks the disk utilization on a server:

CODE: SELECT ALL define service{ hostgroup_name HP-UX_Servers, AIX_Servers_-_NRPE, Linux_Servers service_description UX - Check All Disks _SERVICE_ID 38 use generic-service check_command check_nrpe_all_disks process_perf_data 1 notifications_enabled 1 }

The service runs and changes to a WARNING HARD state. I have an escalation that sends out a single email:

CODE: SELECT ALL define serviceescalation{ servicegroup_name SG_Check_All_Disks contact_groups UNIX_Admin_-_EMAIL first_notification 1 last_notification 1 notification_interval 1440 escalation_period 24x7 escalation_options w }

If the service check goes into a CRITICAL HARD state instead of a WARNING HARD state a helpdesk ticket is created. Here is the escalation that accomplishes that:

CODE: SELECT ALL define serviceescalation{ servicegroup_name SG_Check_All_Disks contact_groups Helpdesk_-_group first_notification 1 last_notification 1 notification_interval 60 escalation_period 24x7 escalation_options c }

In either case my escalation works when the state starts at OK. However, my helpdesk ticket (CRITICAL HARD escalation) does not get triggered when the state goes from WARNING to CRITICAL.

I have found this problem identified on another website:

mail-archive.com/nagios-user … 29862.html

and it seems that it is a known problem. However, until it is fixed how do you get around it? I.E. How do you escalate a single time on a CRITICAL HARD regardless of the state it came from (either OK or WARNING)?

I know that I can set up notifications options to CRITICAL for the helpdesk CONTACT and then set the helpdesk escalation to trigger more than just the first time. However, I would then end up with multiple helpdesk tickets for the same problem.

So, any ideas that you have or any examples on how you resolved this in your own environment would be greatly appreciated.

Thanks