Service escalation notifications not sending

[akepley]

I have setup new host and service escalations for our servers. The host notification is going out just like we expect. The service escalations are not going out at all.

I see this in the log:

Code: Select all

[1407526455] HOST ALERT: server_name;DOWN;HARD;5;CRITICAL - 172.17.16.205: rta nan, lost 100%
[1407526455] HOST NOTIFICATION: akepley;server_name;DOWN;xi_host_notification_handler;CRITICAL - 172.17.16.205: rta nan, lost 100%
[1407526455] HOST NOTIFICATION: dthurston;server_name;DOWN;xi_host_notification_handler;CRITICAL - 172.17.16.205: rta nan, lost 100%
[1407526455] HOST NOTIFICATION: joliver;server_name;DOWN;xi_host_notification_handler;CRITICAL - 172.17.16.205: rta nan, lost 100%
[1407528302] HOST NOTIFICATION: akepley;server_name;DOWN;xi_host_notification_handler;CRITICAL - 172.17.16.205: rta nan, lost 100%
[1407528302] HOST NOTIFICATION: dthurston;server_name;DOWN;xi_host_notification_handler;CRITICAL - 172.17.16.205: rta nan, lost 100%
[1407528302] HOST NOTIFICATION: joliver;server_name;DOWN;xi_host_notification_handler;CRITICAL - 172.17.16.205: rta nan, lost 100%
[1407528302] HOST NOTIFICATION: helpdesk;server_name;DOWN;xi_host_notification_handler;CRITICAL - 172.17.16.205: rta nan, lost 100%

I see the Service Alerts in the logs, but no SERVICE NOTIFICATION logging at all.

Attached are the configs for the service escalation.

[Box293]

I noticed this:

Code: Select all

escalation_options                      c,u,r
escalation_options                      c

However no alerts for services that go into a warning state.

Is it possible that in your tests the services have been going into a warning state and hence everything is working as expected?

[akepley]

Here is a question: If I take the server down, does it ignore service alerts and just notifies of the host alert? I was hoping to get alerts for all for testing, but maybe shutting down the server just makes it not send service alerts. Sorry if this is working as it's suppose to and I didn't realize.

[tmcdonald]

If a host goes down and the services connected to it become warning/critical, Nagios should only notify about the host state. I can't find the exact part in our documentation that says this, and I will update this post when I do.

[Box293]

This post explains how notifications occur and how a server being "down" is defined:

http://support.nagios.com/forum/viewtop ... 10#p103689

[akepley]

Ahh, okay. That clears that up for me. Okay, then everything is working like it's suppose to. Thanks for the clarification.