Notifications going to wrong user

[mkeey]

We have set up NagiosXI to send different notifications to different teams. Example: Warnings go to the support teams and Criticals go to our 24x7x365 help desk team.

Defined users, set up the emails, deployed to the appropriate users but found out today that the help desk team is receiving Warning emails. Not sure how I can check this to find out where the problem lies.

Assumed they would be in the CFG file for each host & service, but there's no distinction between Warn/Crit in the config...

define service {
host_name FAKENAME
service_description Memory Usage
use xiwizard_windowsserver_nsclient_service
check_command check_xi_service_nsclient!!MEMUSE!-w 90 -c 95!!!!!
max_check_attempts 24
check_interval 10
retry_interval 5
check_period xi_timeperiod_24x7
notification_interval 0
notification_period xi_timeperiod_24x7
notifications_enabled 1
contacts serverteam,helpdeskteam
_xiwizard windowsserver
register 1
}

How can I correct this so emails go to the proper people? Do I just load & redeploy the emails on the notifications screen?

Regards,
Martin

[lmiltchev]

There is more than one way of achieving this. If you wanted ALL of the warning notifications to go to the support team, and ALL of the critical notifications to go to the help desk team, you could modify Notification Preferences for each xi user (who is also a contact) on each team to only accept service notifications of certain type.

Example of Notification Preferences for a member of the help desk team:

example01.PNG

If you wanted warning and critical notifications to go to the correct team ONLY FOR THIS SERVICE (Memory Usage on FAKENAME host), then you would need to create two separate services.

Example:

Code: Select all

define service {
host_name FAKENAME
service_description Memory Usage serverteam
use xiwizard_windowsserver_nsclient_service
check_command check_xi_service_nsclient!!MEMUSE!-w 90 -c 95!!!!!
max_check_attempts 24
check_interval 10
retry_interval 5
notification_options        w,
check_period xi_timeperiod_24x7
notification_interval 0
notification_period xi_timeperiod_24x7
notifications_enabled 1
contacts serverteam
_xiwizard windowsserver
register 1
}

define service {
host_name FAKENAME
service_description Memory Usage helpdeskteam
use xiwizard_windowsserver_nsclient_service
check_command check_xi_service_nsclient!!MEMUSE!-w 90 -c 95!!!!!
max_check_attempts 24
check_interval 10
retry_interval 5
notification_options        c,
check_period xi_timeperiod_24x7
notification_interval 0
notification_period xi_timeperiod_24x7
notifications_enabled 1
contacts helpdeskteam
_xiwizard windowsserver
register 1
}

[mkeey]

Thank you for the response and information. It's always a good learning experience to see some details on how things work.

But, I think my message may have been a little misunderstood so please, let me rephrase...

We have all Services setup that if a Warning is received the Email Notification should go to the operating support team(s). So, if it's Redhat, it goes to our Linux Admins. If it's Windows, it goes to our PC Team. If it's CentOS, again to the Linux Team. And so on. Only! The Help Desk is NOT notified at this point. The idea is we "warn" the support teams so they can adjust resources or Nagios thresholds to prevent any Critical issues.

Any Services that has at Critical an Email Notification should go to our Help Desk only. They then call out the problem and assign an incident to the support team(s).

That's how things SHOULD be working.

Today, I found out that on 11/21/18 the Help Desk started receiving Warning Notifications. That should not be!

So, I need to find out where things are messed up and I'm not sure how to go about that. The Notification process, as you pointed out, is somewhat complicated and convoluted. Many places something could get messed up.

Now that you've been given the desired notification process, could you Please advise as to what steps I can take to diagnose where the issue
may be.

Thanks - Martin

[lmiltchev]

According to your config:

define service {
host_name FAKENAME
service_description Memory Usage
use xiwizard_windowsserver_nsclient_service
check_command check_xi_service_nsclient!!MEMUSE!-w 90 -c 95!!!!!
max_check_attempts 24
check_interval 10
retry_interval 5
check_period xi_timeperiod_24x7
notification_interval 0
notification_period xi_timeperiod_24x7
notifications_enabled 1
contacts serverteam,helpdeskteam
_xiwizard windowsserver
register 1
}

ALL notifications (all types, e.g. Warning, Critical, etc.) will be sent to both contacts - serverteam and helpdeskteam.

Can you log in the web UI as helpdeskteam, and show us a screenshot of the notification preferences page?

[mkeey]

Ok, I can log onto the Web GUI with the HelpDesk account, but it's not an Admin account so the Admin/Configure pages nor the "my account settings" page display.

[mkeey]

Let me maybe shed a little more light on this. We have two notification templates.

1st = Warning

Warning.jpg

2nd = Critical

Critical.jpg

We then use the deploy function to assign them to users. So, it makes sense that both users are listed in the CFG file because the Service Warning goes to one user and the Service/Host Critical goes to another user.

[mkeey]

Can I just redeploy the Notification Templates and that will correct the issue?

[benjaminsmith]

These templates look correct. The way you have them set up, notifications *should* go to correct contacts, provided they were deployed properly. You can double check if in the contacts config you have the following:

Code: Select all

define contact {
    contact_name                     serverteam
    ...
    service_notification_options     w
    ...
}

and

Code: Select all

define contact {
    contact_name                     helpdeskteam
    ...
    service_notification_options     c
    ...
}

You can also try redeploying the templates to proper contacts to see if this is going to solve the issue.

FYI, I tested a similar scenario with the same templates, and I can verify that this setup is working as intended. See below.

[mkeey]

This is cool!! Now we're getting somewhere. I'm slowly learning how to diagnose the problem. Thank you!

Anyway, I went to CCM screen and clicked on Contacts and got the list of users. Brought up the Contact Config file and extracted the users in question. Here is a side-by-side comparison. You can see the differences and why the user is getting Warning Notifications. Funny thing is, they are also getting Critical's but there isn't a "C" in addition to the "W"...


So, then I looked at the actual GUI Contact screen and can see that the wrong boxes are activated...


How can I update either the Config file or the Contact form? Never did this, so I'm hesitant to screw anything up. Can I just click on the appropriate boxes? Or, should I redeploy from the Notification Template screen?

Thanks - Martin

[lmiltchev]

Let's step back. In your first post, you showed us the following config:

define service {
host_name FAKENAME
service_description Memory Usage
use xiwizard_windowsserver_nsclient_service
check_command check_xi_service_nsclient!!MEMUSE!-w 90 -c 95!!!!!
max_check_attempts 24
check_interval 10
retry_interval 5
check_period xi_timeperiod_24x7
notification_interval 0
notification_period xi_timeperiod_24x7
notifications_enabled 1
contacts serverteam,helpdeskteam
_xiwizard windowsserver
register 1
}

The idea as to configure serverteam contact to receive only warning notifications, and helpdeskteam contact to receive criticals, correct?

In your last screenshot, you are showing us a completely different contact - prodservices... What this contact has to do with what you wanted to accomplish (per your first post)?

I believe you need to apply your FTL_UserWarning template to the serverteam contact and FTL_UserCritical template to the helpdeskteam contact.