Introduction
OpenSearch was created by AWS in 2021 when Elastic changed Elasticsearch from Apache-2.0 to a non-open-source license. Forked from Elasticsearch 7.10.2, OpenSearch provides the same core search and analytics capabilities with a commitment to true open-source licensing (Apache-2.0).
With over 13,000 GitHub stars and backed by AWS, OpenSearch has grown beyond a fork — it now includes unique features like neural search, conversational search, and observability tools not found in the original Elasticsearch.
What OpenSearch Does
OpenSearch indexes and searches structured and unstructured data at scale. It is wire-compatible with Elasticsearch 7.x, meaning existing Elasticsearch clients, tools, and integrations work with OpenSearch. The suite includes OpenSearch (search engine) and OpenSearch Dashboards (visualization, equivalent to Kibana).
Architecture Overview
[OpenSearch Suite]
|
+-------+-------+
| |
[OpenSearch] [OpenSearch Dashboards]
Search engine Visualization & UI
(Elasticsearch (Kibana fork)
fork) Alerting, Notebooks
|
[Plugins (built-in)]
Security (free!)
Alerting
Anomaly Detection
SQL/PPL queries
Neural Search (AI)
Observability
Index Management
|
[Compatible with]
Logstash, Fluentd,
Beats, Data PrepperSelf-Hosting & Configuration
# docker-compose.yml
version: "3.8"
services:
opensearch:
image: opensearchproject/opensearch:2.15.0
environment:
- cluster.name=my-cluster
- discovery.type=single-node
- bootstrap.memory_lock=true
- "OPENSEARCH_JAVA_OPTS=-Xms2g -Xmx2g"
ulimits:
memlock: { soft: -1, hard: -1 }
ports:
- 9200:9200
volumes:
- os-data:/usr/share/opensearch/data
dashboards:
image: opensearchproject/opensearch-dashboards:2.15.0
ports:
- 5601:5601
environment:
OPENSEARCH_HOSTS: "[\"https://opensearch:9200\"]"
volumes:
os-data:Key Features
- Apache-2.0 License — truly open source with no feature gates
- Security Built-in — free authentication, RBAC, encryption, and audit logs
- Neural Search — AI-powered semantic search using ML models
- SQL/PPL — query with SQL or Piped Processing Language
- Alerting — built-in alerting with notifications (Slack, email, webhook)
- Anomaly Detection — ML-powered anomaly detection on time-series data
- Observability — trace analytics, log correlation, and metrics
- Elasticsearch Compatible — works with existing ES 7.x clients
Comparison with Similar Tools
| Feature | OpenSearch | Elasticsearch | Meilisearch | Typesense |
|---|---|---|---|---|
| License | Apache-2.0 | SSPL/Elastic | MIT | GPL-3.0 |
| Security (free) | Yes | Paid | Basic | Basic |
| Alerting (free) | Yes | Paid | No | No |
| ML/AI Search | Neural search | Paid | AI-powered | Semantic |
| Scale | Petabytes | Petabytes | Single-node | Cluster |
| AWS Managed | Amazon OpenSearch | Elastic Cloud | Cloud | Cloud |
| Best For | OSS enterprise search | Latest features | Simple apps | Simple apps |
FAQ
Q: Can I migrate from Elasticsearch to OpenSearch? A: Yes. OpenSearch is wire-compatible with Elasticsearch 7.x. Most clients, tools, and dashboards work with minimal changes. AWS provides migration documentation.
Q: Is OpenSearch just an Elasticsearch fork? A: It started as a fork but has diverged significantly. OpenSearch has added neural search, conversational search, flow framework, and other features not in Elasticsearch.
Q: Why is OpenSearch free security important? A: Elasticsearch requires a paid license for basic security (authentication, encryption). OpenSearch includes comprehensive security for free — critical for production deployments.
Q: Should I choose OpenSearch or Elasticsearch? A: OpenSearch for true open-source licensing, free security features, and AWS integration. Elasticsearch for the latest search innovations and Elastic Cloud ecosystem.
Sources
- GitHub: https://github.com/opensearch-project/OpenSearch
- Documentation: https://opensearch.org/docs
- Created by AWS, community-maintained
- License: Apache-2.0