Trap for loss of connectivity to managed Mobility controller on MM 8.3.0.7

[fbarta]

Hello,

I am currently working on integrating my first Aruba network against an existing NMS platform. I've reviewed the list of configurable SNMP traps on the MM under "show snmp trap-list", but I do not see any events that jump out to me as indicative that one of my managed MCs has gone offline.

Is the MM capable of sending a SNMP trap when connectivity to one of the managed MCs is lost? Which event should I be looking at?

[jgoff · Accepted solution]

there is, unfortunately, no trap for this :| 

There are alternatives you can cobble together depending if you have some sort of SIEM pipeline (things like logstash or splunk etc) by looking for specific logs that come out of the controller syslog when clustering breaks, or, you can write some basic tools to interrogate the controller API to get the cluster state and said tool can alert your NMS.

This can also be done via some hacking about in Aiwave to look for the specific logs that come in from clustermgr when the clustering breaks, and that can forward something off to your NMS.

[jgoff · Accepted solution]

there is, unfortunately, no trap for this :| 

There are alternatives you can cobble together depending if you have some sort of SIEM pipeline (things like logstash or splunk etc) by looking for specific logs that come out of the controller syslog when clustering breaks, or, you can write some basic tools to interrogate the controller API to get the cluster state and said tool can alert your NMS.

This can also be done via some hacking about in Aiwave to look for the specific logs that come in from clustermgr when the clustering breaks, and that can forward something off to your NMS.

[fbarta]

Thank you for the suggestions! Will definitely research the logs on cluster state changes for notification of when a MC goes fully offline.

My concern is that I've experienced failure scenarios with other vendors in the past where a device may be online and otherwise functioning normally, but the communication to the central controller is not working for one reason or another (Unexpected firewall rule changes, software defects, etc). I'm probably getting off into edge case territory, but it would be really nice if the MM itself had the ability to notify on loss of connectivity to managed controllers.