Azure SQL Database Data Sync Retirement: Migration Scenarios and Recommended Alternatives

Azure SQL Data Sync, relying on triggers and metadata tables, is being retired due to its complexity and performance limitations. Microsoft recommends migrating to more scalable services like Azure Data Factory (ADF) or event-driven replication. This article provides a sample scenario consolidating data from multiple Azure SQL Databases into one, highlighting the migration process. Before deciding on a replacement, consider factors like latency, write patterns, and schema compatibility. The current setup with Data Sync is illustrated, contrasted with the target state using ADF, which offers advantages like parallel ingestion. Incremental replication using ADF, Change Tracking, or watermark columns minimizes data loss, achieving a Recovery Point Objective of 1-5 minutes. The migration approach involves assessments, initial seeding, and incremental replication, with a recommended DEV to PROD rollout. Azure Data Factory is the primary recommended replacement due to its full support, scalability, and built-in monitoring. Other options like SQL Transactional Replication and Azure SQL Managed Instance Link are discussed, although less favored. The migration minimizes service downtime and provides predictable performance impact and built-in observability via ADF monitoring. This migration presents the opportunity for cleaner architectures and easier production operations aligned with the Azure platform. If you're using Data Sync, initiate your migration planning now.