We currently have one server running the indexer and search head. We're looking into adding a second indexer on a different host to take advantage of license pools. The new indexer would be set up to where only a subset of our clients would be pointing to it as we've had issues with another group consuming too much of our license. If we take that approach, will the data be replicated between both indexers to where we can take advantage of additional search resources or do indexers only keep they data being sent to them by the forwarders configured to use them?
By default indexers do not automatically replicate data. You can enable this behaviour though by setting up index clusters. The docs have excellent explanations on how it works in detail, so I'll just refer to those: http://docs.splunk.com/Documentation/Splunk/6.1.2/Indexer/Aboutclusters
Ok, so I'm a little confused on the data the forwarders are sending. Do you have to load balance the forwarders between all indexers or can I point forwarders A,B,C to indexer 1 and forwarders D,E,F to indexer 2 and still have replication work correctly?
The indexer that receives the data from the forwarder will be the primary bucket for that data.
If you configure clustering/replication, the primary bucket will be copied to other indexers to meet whatever replication factor has been configured.
In a single site setup, any search will prefer to use the primary bucket for returning results if it is available.