Search and vector store
Amazon OpenSearch Service
Managed OpenSearch clusters for search, log analytics, application monitoring, and vector search with provisioned capacity control.
Key points
- Provides a managed OpenSearch search and analytics engine based on the Elasticsearch fork.
- Supports log analytics, full-text search, application monitoring, and vector search workloads.
- Uses a provisioned cluster model where teams select and manage capacity.
- Fits steady workloads that need explicit capacity, sizing, and operational control.
When to use it
- Run predictable large-scale log analytics or search workloads with cluster capacity control.
- Build full-text search and observability applications on managed OpenSearch clusters.
- Host vector search when a team wants provisioned cluster management rather than serverless capacity.
Exam tips
- OpenSearch Service means provisioned managed clusters with capacity planning.
- Do not pick provisioned OpenSearch Service as the default Bedrock Knowledge Bases vector store; that default is OpenSearch Serverless.
- Choose it when the question emphasizes steady log analytics, full-text search, or capacity control.