엘라스틱서치 핵심 개념

Let's master Elasticsearch's core concepts: clusters, nodes, indices, documents, and the shards and replicas that make it scale.

An Elasticsearch cluster is one or more nodes that together hold all your data and serve indexing and search as a single, unified system.

A node is one server in your cluster, running an Elasticsearch instance and storing part of your data. Each has a unique name and can take on different roles.

Nodes sharing a cluster name discover each other and form a cluster. One becomes the master node, managing cluster-wide changes; the rest hold data.

An index is a logical grouping of similar documents, like a database in the relational world — say a products index and a customers index.

A document is the basic indexed unit: a single JSON record with a unique ID inside its index. The example shows one product document.

A mapping defines how each field is stored and indexed. It shapes how you can search — text for full-text, float for price, and so on.

Elasticsearch splits an index into shards — self-contained sub-indexes spread across nodes for horizontal scaling and parallel processing.

A replica shard is a copy of a primary shard. If a node fails, a replica takes over with no data loss — and it boosts search throughput too. 🛡️

Putting it together: a cluster of nodes hosts shards, shards are slices of an index, an index holds documents, and replicas add redundancy.

Recap: clusters, nodes, indices, documents, mappings, shards, and replicas are the building blocks of Elasticsearch. Next, you'll spin one up.