vROps multinode SNMP gotcha
- 8 nodes if you are on versions 6.0.x
- 16 nodes if you are on version 6.1
Why does this matter if you plan to use SNMP with your vROps cluster?
Quite simple really, what I discovered today is that even though you have a master node SNMP traps can and will be sent from all of the nodes to the destination you have configured.
Why is this a problem?
If you do not configure your SNMP trap receiver (lets say a vRO instance) to receive SNMP traps from all of the possible node IP’s (or FQDNs) you will miss the traps sent from the other nodes in your vROps cluster.
That’s a pretty big deal if you are reliant on those traps for event/alert notification!
As far as I can determine traps will be sent from the nodes that have been configured as collectors for any given monitored source. But of that I am not 100% sure.
So there we have it, if you plan to use SNMP or already have it configured, make sure you configure your SNMP receiver to account for all of the nodes in your vROps cluster.