Overview
Chongqing Municipal Medical Security Bureau
Country: China
Industry: Government / Healthcare
Solution: i2Active for MySQL (Agentless Synchronization)
To support the unified development of Chongqing’s medical insurance data-sharing platform, Info2soft implemented an agentless synchronization solution that enables secure, high-performance replication across districts. The project ensured accurate, real-time sharing of healthcare data across heterogeneous databases while maintaining production stability.
Customer Challenge
- Establish a unified citywide medical insurance data-sharing platform covering multiple districts.
- Synchronize data across heterogeneous MySQL and HotDB environments without installing agents on production servers.
- Automatically generate target tables and track change operations with timestamps.
- Resolve primary key conflicts intelligently while maintaining the most recent valid record.
- Prevent data inconsistencies between source and target systems under high-volume transactions.
I2 Solution
- Agentless Synchronization Architecture: Deployed i2Active for MySQL with no software installation on production servers.
- Auto Table Creation: Automatically generated target-side tables with operation type and timestamp fields for auditability.
- DELETE-to-INSERT Conversion: Transformed DELETE operations into INSERTs tagged with a “D” flag to retain historical traceability.
- Conflict Handling: Applied configurable load policies to retain the latest version of data on key conflicts.
- Incremental Synchronization: Detected duplicate rows and performed delete-reinsert operations for full consistency.
- Dedicated Sync Servers: Two synchronization hosts extracted logs remotely from HotDB and replicated them to the target MySQL databases through Info2soft nodes.
Solution Effect
- Unified Management Platform: All synchronization tasks and rules managed centrally in the i2UP visual interface, reducing DBA workload.
- Zero Impact on Production: Achieved fully agentless synchronization through log-based capture without any resource occupation on source servers.
- High-Efficiency Labeling Completed: Successfully implemented record-level labeling for over 10,000 middleware tables and nearly 2,000 operational tables, meeting national compliance and integration standards.