Elastic Docs: Elasticsearch

What Elasticsearch is

Elasticsearch is a search and analytics engine. You send it JSON documents, it builds an inverted index (a map from every word to the documents that contain it), and then it answers full-text queries in milliseconds, ranked by relevance. Elastic describes its platform as a search, analytics and AI platform, with ready-made solutions for observability (logs and metrics) and security built on the same engine.

What the docs cover

• Get started: the core ideas (indices, documents, mappings), the deployment choices and a local setup so you can try it on your own laptop.
• Search and queries: match and multi-match queries, filters, aggregations for facets such as "brand" or "price range", highlighting and pagination.
• Text analysis: analysers, tokenisers and language support, which decide how "Running Shoes" matches "run shoe".
• Vectors and hybrid search: dense vector fields and combining semantic and keyword search.
• Operations: index lifecycle, snapshots and upgrades for when you run it yourself.

Is it worth learning?

If your app needs search that feels like a real search bar (typo tolerance, ranking, facets, highlighting across thousands or millions of items), yes. It is also widely used for logs, so the skill transfers to DevOps work. If you only need simple search over a few thousand rows, try PostgreSQL full-text search first: one less system to run.

Licence and the OpenSearch fork

In 2021 Elastic moved Elasticsearch away from Apache 2.0, and AWS forked it as OpenSearch, now run under the Linux Foundation. In September 2024 Elastic added the OSI-approved AGPLv3 as an option alongside SSPL and the Elastic License, so you can choose. OpenSearch is the engine behind Amazon OpenSearch Service. The core ideas (indices, mappings, analysers, query DSL) are close enough that learning one teaches you most of the other.

Prerequisites and time

Comfort with JSON and HTTP requests is enough. Expect an evening to index a dataset and run your first queries, and a weekend to understand analysers, scoring and aggregations properly.

Free or paid?

The self-managed engine is free to download and run. Elastic Cloud is a paid managed service with a free trial (14 days as of October 2026).

Common questions

Should Elasticsearch be my main database? No. Keep your source of truth in a database such as PostgreSQL and copy what needs searching into the index.

Best for: backend and full-stack developers adding serious search to an app, and anyone working with logs.