Adapting Pagination from SQL Server to Amazon DynamoDB for .NET Applications

This article explains how to adapt pagination patterns from SQL Server to Amazon DynamoDB for .NET applications, comparing offset-based and token-based pagination models.

• SQL Server uses offset-based pagination with row positions; DynamoDB uses token-based continuation with LastEvaluatedKey
• Page number navigation requires sequential scanning to simulate jumps; costs increase linearly with page depth
• Previous/next navigation maps directly to DynamoDB's model; maintains consistent per-page costs without filters
• Infinite scrolling accepts variable batch sizes; reduces API calls and simplifies filter handling
• FilterExpression applies after items are read, affecting page size consistency and RCU consumption
• Replace page numbers with serialized continuation tokens; remove total count properties for cost efficiency
• Choose pattern based on UI requirements, navigation depth, filter selectivity, and cost constraints

Migrating pagination to DynamoDB requires aligning design with token-based architecture and actual access patterns rather than forcing offset-based patterns onto distributed systems.