An indexer that is used for search processing in a parallel reduce search process. During a parallel reduce search, a subset of the indexers in a Splunk Enterprise deployment process a portion of a search before they push their results to the search head, where final processing and result aggregation takes place. This parallelization of work that otherwise would be done entirely on the search head can result in faster completion times for searches over large high-cardinality datasets.
By default, when you initiate a parallel reduce search job, the Splunk platform randomly selects the subset of indexers that process the search as intermediate reducers. Alternatively, you can designate a set of indexers to be dedicated intermediate reducers. That set of indexers will always serve as intermediate reducers when you run parallel reduce search jobs.
For more information
In Distributed Search: