Prime Infrastructure 3.2 unable to create notification policy

[steiostb1]

After upgrading to 3.2 I am not able to setup Prime to send emails when APs is down eg. In 3.2 they have removed the email notification field in the alarm view. Instead you have to create "Notfication Policies". I have tried multiple times to create this, but this error message always pop-ups: "Unable to Create Policy. Please try again later". Now i have no email sent out when APs are now..

Any tips here? This could be a major bug.

Might be this bug i am experiencing

CSCvd67569
Getting error message when editing the notification destination with virtual domain value.

Regards

[steiostb1]

Is this fixed now?

[Thomas Hausruckinger]

Hm, works here.

Have you created "Notification Destinations"?

I think there have to be at least one "Notification Destination"under Administration / Settings / System Settings/ Mail and Notification

[steiostb1]

Yes, i have created notification. I am currently using PI 3.3 and I still have the same problem. I receive emails for saved reports.

I have verified this so far:

• Notification destination (Under Administration - settings - system settings - Mail and Notifcation) contains a contact for each policy. 
• Alarm notification Policies (Under Administration - settings - system settings - Alarms and Events) contains policies and they have been checked and looks correct. I have made a policy to send a email whenever an AP disassociated from the controller. 
• I have run the Test to test mail configuration from Mail server configuration (Under Administration - settings - system settings - Mail and Notifcation). I received this email.

I do still not get email when an AP disassociated from the controller

Any thoughts?

Thanks

[steiostb1]

More troubleshooting:

I have turned logging to trace-Level and found this in ncs_nb.log:

2018-01-26 11:24:03,882 [notifPolicy-queueRecipients-1] DEBUG services - EmailService: Pushing alarm into email queue as no suppression for alarm, alarmId LoadProfile_LradIf!c0:62:6b:24:c3:b0!0
2018-01-26 11:24:03,890 [notifPolicy-queueRecipients-1] DEBUG services - processAlarmForwardedFromFaultPolicy: processAlarmForNotificationNew done: LoadProfile_LradIf!c0:62:6b:24:c3:b0!0
2018-01-26 11:24:03,890 [notifPolicy-queueRecipients-1] DEBUG services - sendToEmailQueue: Added into queue: emailReceivers: [steiostb]
2018-01-26 11:24:03,892 [notificationPolicyListener-Thread] DEBUG services - Applying the alarm to the policies..RogueAp!00:1d:c9:07:36:49
2018-01-26 11:24:03,893 [notifPolicy-queueRecipients-0] DEBUG services - Not Pushing alert into email queue as emailNotificationFlag is false: alarmId LoadProfile_LradIf!c0:62:6b:24:c3:b0!0

Not Pushing alert into email queue as emailNotificationFlag is false - what can this mean? Is there some settings i have missed?

Thanks

[sdugravot]

Hello,

Same problem here with Prime 3.3 (unable to create or delete notification policy)

Have you solved the problem?

Stephane.

[snoc]

Hey,
same issue here with prime 3.4 (just migrated form 3.2)
From the migration we had a lot of notification policies, (per user, and device category one policy)
as I wanted to clean this up, I got the same error.

[steiostb1]

I manage to solve this by deleting the notification policy (which was imported during the migration) and create a new notfication policy. It seems like the migration from 3.1/3.2 fails for notification policy.

[Kyle Nielsen]

I'm having the same issue, and currently have a TAC case open. 

I also tried all of the same steps as steiostb1 and thank you for posting the ncs_nb.log, after checking there I was seeing "Not Pushing alert into email queue as emailNotificationFlag is Suppressed"  but there is no setting in place to suppress the the alarm. Opened a TAC case and the engineer saw a java error being thrown when the alert should have been triggered... 

I will post and update when the case is resolved. 

[Kyle Nielsen]

Update... had the issue escalated with TAC and PSTS, they found another case with the same issues after upgrading from 3.1 and the issue is in both 3.2 and 3.3,  the "fix" is to upgrade to 3.4. I plan on doing that tomorrow, I will update again with results.