NOTE
Elasticsearch Architecture
Cluster nodes, indices, primary/replica shards, coordination, routing, distributed search, and recovery.
This is a historical learning note and may contain outdated or incomplete understanding.
An Elasticsearch index is partitioned into primary shards; replica copies provide redundancy and additional read capacity. Shards are Lucene indexes distributed across cluster nodes.
Any suitable node can coordinate a client request. Document routing identifies the target shard for point writes/gets; searches fan out to relevant shard copies and combine partial results.
Cluster-manager/master-eligible coordination maintains cluster metadata/allocation decisions, while data nodes execute shard work. Node roles and coordination internals have changed across releases, so use role terminology/documentation matching the deployed version.