Thread pools:
- bulk: to support the bulk api which let you perform many index/delete operations in a single API call
- flush: to flush indices in memory data to the disk
- generic: to support background node discovery
- get: to support the Get API which let you retrieve a document from an index based on its id
- index: to support index/delete operations
- management: to support cluster management
- force_merge: to support the merge of segments a Lucene index holds in each shard. The merge operation merge the segments to reduce the number of segments.
- percolate: to support percolate query (you store search queries and throw documents at it)
- refresh: to support the force of indices refresh. Refresh is internally scheduled periodically but you can force it.
- search: to support search and count queries
- snapshot: to support the creation of indices or cluster backup on remote repository
- suggest: to support search suggestions: suggest related searches based on the current one.
- warmer: to support indices warm up: runs registered searches to warm up the index. This is deprecated since 2.3.0 thanks to disk-based norms and doc values.