OpenSearch

What it is

OpenSearch is a fork of Elasticsearch 7.10.2 and Kibana, created after Elastic relicensed those projects away from open source in early 2021. It's a distributed search and analytics engine used for log analytics, full-text search, application monitoring, and security analytics (via OpenSearch Dashboards and plugins like Security Analytics and Observability). Under the hood: Lucene-based indexing, REST APIs, a query DSL, and horizontal scaling across nodes/shards — functionally familiar to anyone who's used Elasticsearch.

Who builds it and why

Originally spun up by AWS as a direct response to Elastic's license change, giving companies running Elasticsearch-based products (including AWS's own managed Elasticsearch Service) a genuinely open alternative. Governance has since moved to the OpenSearch Software Foundation under the Linux Foundation, with contributions from AWS, SAP, Uber, Adobe, and a broad set of independent contributors — 2,178 on GitHub at last count. The motivation is straightforward: keep a production-grade search/analytics stack under Apache 2.0 permanently, without a vendor able to unilaterally change the license again.

Production readiness signal

13,231 GitHub stars and active commits as recently as July 2026 signal a project with real ongoing engineering investment, not a dormant fork. 2,178 contributors is a large, diverse base — reduces bus-factor risk. Latest release is 3.7.0, indicating steady major/minor version progression rather than stagnation at a 1.x fork point. It's not a CNCF project (maturity status not publicly available, and it isn't hosted there — it lives under the Linux Foundation via the OpenSearch Software Foundation). AWS runs OpenSearch as a managed service (Amazon OpenSearch Service) at significant scale, which is a strong real-world production signal independent of GitHub metrics.

Who should use this

• Teams already running Elasticsearch 7.x who need a license-compatible upgrade path without vendor lock-in.
• Organizations building log analytics or SIEM-style pipelines who want Apache 2.0 licensing for redistribution or embedding in commercial products.
• AWS-centric shops that want a managed offering (Amazon OpenSearch Service) without operating the cluster themselves.
• Teams needing full-text search plus visualization (Dashboards) in one open-source stack, without paying for Elastic's commercial tiers.

Who should NOT use this

• Teams wanting the latest Elasticsearch/Kibana features — Elastic's proprietary tree has diverged and often ships new capabilities (certain ML features, some security integrations) first or exclusively.
• Shops needing a small, single-binary, low-ops search tool — OpenSearch is a JVM-based distributed system with real operational overhead (heap sizing, shard management, cluster tuning). If you just need embedded search, look at something like Meilisearch or Typesense instead.
• Organizations wanting a single vendor accountable end-to-end with enterprise SLAs baked into the OSS project itself — you'll need AWS or a third party (e.g., a support vendor) for that, the OSS project alone doesn't provide it.
• Teams with hard requirements for feature parity with Elastic's commercial X-Pack security/ML modules — OpenSearch's Security and ML plugins are separately maintained and don't guarantee 1:1 parity.

Alternatives

• Elasticsearch — the original, now dual-licensed (SSPL/Elastic License + a source-available option); more proprietary features but license terms restrict certain commercial uses.
• Meilisearch — lightweight, developer-friendly search engine better suited for smaller-scale, low-latency use cases than large-scale log analytics.
• Apache Solr — another Lucene-based search platform, more mature in some enterprise search contexts but with a smaller modern observability ecosystem than OpenSearch.

Pricing

Fully open source under Apache-2.0 — no license fees, no feature gating. You pay only for infrastructure and, optionally, managed hosting (e.g., Amazon OpenSearch Service, or third-party providers) or paid support contracts if you want them.