How to migrate from Amazon CloudSearch to Amazon OpenSearch Serverless

This article provides a comprehensive guide for migrating from Amazon CloudSearch to Amazon OpenSearch Serverless, enabling access to modern search capabilities like semantic search, RAG, and agentic search with automatic scaling.

• Assess CloudSearch configuration including instance types, field definitions, analyzers, and API versions before migration
• Model data using explicit OpenSearch index mappings with strict validation to prevent schema drift
• Map CloudSearch field types to OpenSearch equivalents (text to text, literal to keyword, latlon to geo_point)
• Convert source documents from CloudSearch SDF format to OpenSearch JSON format and store in Amazon S3
• Use Amazon OpenSearch Ingestion pipelines to bulk-load converted documents from S3
• Translate CloudSearch URL-based queries to OpenSearch query DSL with compound queries and boosting
• Configure layered security using encryption, network, and data access policies
• Validate document counts, query results, ranking, and latency before production cutover
• Monitor OpenSearch Compute Unit (OCU) consumption and set capacity limits for cost management

OpenSearch Serverless provides hands-off operations with independent scaling of indexing and search compute, pay-per-use pricing, and richer query capabilities than CloudSearch.