Is the JMS Messaging Modular Input Cluster aware?
In other words, can it be managed from the Deployer server and distributed to multiple search heads in a cluster ?
OR
Does this live on a standalone instance?
The basic question I am asking is, does this support HA / failover?
In a distributed environment, the JMS App should be installed on 1 or more forwarders. The data will then be indexed into your indexer cluster and searchable from your search head cluster.
Just to be clear, can you configure more that 1 forwarder with the same config?
The way I read the documentation was that the JMS App pulled data from the event source.
If 2 ore more forwarders have the same config, then data would/could be duplicated.
HA is a requirement for my specific use case.
Yes you can if pulling data from JMS queues.Data does not get duplicated because you are dequeing messages (unlike with JMS topics.)
This is a standard approach for achieving horizontal scalability.