Search Head Pool and Clustering

[kenmcgarrahan]

Are cluster configuration and search head pooling compatible? In other words, can one configure a master node that references a search head pool? The "Managing Indexes and Clusters" document implies it is possible:

If you want to use some of the more advanced features of distributed search, such as search head pooling or mounted bundles, you must edit distsearch.conf on the search head. For detailed information on search heads, including instructions on how to do advanced configuration, read the Distributed Search manual. That chapter focuses on non-clustered environments, but it also provides correct information for configuring advanced features on clustered search heads.

However, the Distributed Search document makes no reference to that configuration.
Does this imply that the master node configuration will reference the search heads in the pool individually, and that the search heads participation in the "pool" is governed by their individual configuration?

[martin_mueller · Accepted solution]

I have set up pooled search heads over clustered indexers for customers on both Splunk 5 and Splunk 6, so those two features must be compatible.

Each search head is added to the cluster under the search head role, and each search head is added to the search head pool as you normally would. The master node doesn't reference search heads at all, regardless of their pooling state. See http://docs.splunk.com/Documentation/Splunk/6.0.3/Indexer/Enablethesearchhead#Add_search_heads_to_a_... for reference.