When testing in the multi-node AWS environment, I noticed that each node had alarms logged in the log directory. This should be the case, all alarms should be processed on the Active Parent OAM Module, which is the module where ProcMgr is running.
When alarms was processed as part of an SNMP-TRAP, it was correct processed on the Active Parent OAM Module because that is where the snmptrapd was running.
Due to the changes to go away from issue snmptraps and just issue alarms, this logic was lost causing the alarms to be process on each local node.
A change is need to have the alarms routed to Procmgr and it would process them. This will cause the processing on the Active Parent OAM Module, like they should do doing