NAS - Duplicated Alarm

[jm_work]

 

Hello CA Guys,

 

We have in our environment, 3 levels of layer between the Hubs.

 

Brazil (Core NMS - Tier I)

Argentina (South Cone - Tier II)

Peru (Customer - Tier III)

 

The customer from Peru is monitored via SNMPCollector probe, so it has local NAS. In the meantime, we are seeing duplicate alarms.

 

As shown in the example attached.

 

 

 

Could you please help us?

 

The alarm is duplicated and processed by different NAS Tiers. [ Here are some examples and also the design of our architecture.

 

 

Would it be possible to review our settings to remove duplicate alarms?

 

CAUIM 8.5.1

NAS 8.56

 

Regards,

Jean Gomes

[Kathy Maguire]

Are there any postroute queues defined between any of these three hubs with alarm or * in the subject?  If so, this is where there duplicate alarms are coming from.  When you have multiple nas probes configured with alarm forwarding, you need to make sure that no alarm messages are exchanged between the hubs where these nas probes are deployed otherwise the alarm message will be processed twice by one or more of the nas probes - once from the downstream hub postroute queue and a second time from the downstream nas probe forwarding the alarms.

[jm_work]

Kathy,

 

Thanks for your help.

 

Just confirming,

 

Hub (tier-II or tier-III) that owns the NAS probe, should not have the queue with subject *;
Hub (tier-II or tier-III) that does not have the NAS probe, can have the queue with subject *;

 

Is correct?

[Kathy Maguire]

Jean -

 

Yes, that is correct.  Hubs that do not have a nas probe installed can have queues defined with subject *.  Hubs with nas probes installed need to have forwarding configured in the nas probe and cannot have queues defined in the hub with subject *.

[Kathy Maguire]

The postroute queues are defined in the hub probes.  You need to check for attach/get queues configured on the Queues tab on the tier1, tier2, and tier3 hubs where you have the nas probes deployed.  None of the attach queues that send messages upstream to the tier1 hub from the tier2 and tier3 hubs should be configured with alarm or * in the subject.

[Yu Ishitani]

Take a look at this. A known issue with asterisk (*).

https://comm.support.ca.com/kb/duplicated-alarm-coming-from-celerra-clariion-probe/kb000043732

 

[Yu Ishitani]

As Kathy mentioned, if you decide to use NAS probe based alarm forwarding, "alarm" subject (in downstream hub) shall not be sent out to upstream hubs via hub queue.