'UP' alert with Unknown, warning or critical alert

[pnewlon]

>> Can you post the email notification messages for the XI users 'nagiosadmin' and another user in the DMB Group? These messages are customized per-user, so you can access them from Configure->My Account Settings->Notification Messages.

I am 'nagiosadmin' and I have a contact 'philnewlon'. I have notifications turned off to user 'nagiosadmin' so that I don't get confused as to whether a message has been sent to the user nagiosadmin or the contact philnewlon. Contact philnewlon is in the contact group 'dmb group'.

>> Your host and service notification period for each XI users needs to be <username>_notification_times in order to work correctly.

I read that doc - there is no 'user <-> contact' correlation in this case.

>> Do the timestamps on the contradicting emails appear to be around the same time, or could there be some issue with the email server getting backed up?

They are around the same time. To eliminate the possibility that there is an email backup as you suggest, I have been using tcpdump watching port 25 so I know exactly when/if a message is spewed out.

I think I have figured out why all messages quit flowing at all for a while - I tried to use the xi_host|service_notification_handler instead of notify_host|service_by_email without a user/contact relationship.

[mguthrie]

define contact {
contact_name philnewlon
alias Phil Newlon
contactgroups DMB Group
host_notifications_enabled 1
service_notifications_enabled 1
host_notification_period 24x7
service_notification_period 24x7
host_notification_options d,u
service_notification_options w,c
host_notification_commands xi_host_notification_handler
service_notification_commands xi_service_notification_handler

You may have changed this since sending me the config snapshot, but at the moment your user needs to be using the "philnewlon_notification_time" for the above timeperiods. Not sure if that's causing an issue or not.

I think I have figured out why all messages quit flowing at all for a while - I tried to use the xi_host|service_notification_handler instead of notify_host|service_by_email without a user/contact relationship.

Yeah, you can add those commands in addition to the xi notification handlers, but if you replace them it will cause problems.

This issue is somewhat dependent on what's in your email inbox, so keep us up to date if anything changes.

[pnewlon]

>> You may have changed this since sending me the config snapshot, but at the moment your user needs to be using the "philnewlon_notification_time" for the above timeperiods. Not sure if that's causing an issue or not.

There is no user 'philnewlon' there is only a contact 'philnewlon'. Does this still hold true in this case? Is it a REQUIREMENT that there be a user defined for every contact??

[mguthrie]

In Nagios XI, I generally recommend creating an XI user for each contact if there are problems related to notifications.

[pnewlon]

deleted contact 'philnewlon', created user 'philnewlon' and allowed the system to create the contact. configured user and contact and now I am getting the 'down' notification but not the 'up'. It's tough to hit a moving target .......

[mguthrie]

Sorry to bug you for a config again, but can you email us the new configuration snapshot with the new XI user? If I can't find the answer in the configs we'll set up a remote session and see what we can figure out.