Project Background
China Mobile Hebei supports large provincial communication services, and its O-domain systems carry one of the highest data volumes and business pressures among provincial branches.
The localization program focused on replacing the database layer across more than 30 business systems while protecting production continuity through an extremely large and complex migration.
Challenges and Requirements
- The migration involved 124.3 TB of Oracle data, 8.05 million lines of PL/SQL code, and 27.28 million database objects.
- Business systems were tightly coupled and contained cross-calls, which meant a poor migration design could trigger production failures.
- The cluster environment had many nodes, creating strong requirements for easy operations, rich monitoring, and safe maintenance.
- The project had to meet carrier governance requirements while ensuring zero data loss and stable cutover.
Implementation Approach
Domestic infrastructure stack
The solution combined Kunpeng chips, BC-Linux, Baolande middleware, Highgo Database Management System V9, and Highgo cluster software as the production baseline.
Deep Oracle-compatibility migration
Highgo used Oracle-compatible mode to carry PL/SQL objects and most source syntax, while upgrading only a limited set of uncommon syntax and packages.
Unified cluster management and joint cutover
The platform relied on Highgo’s unified management console for graphical deployment and operations, plus on-site migration experts for joint testing across O, B, and M domains.
Project Highlights
- Delivered sub-second response for front-end queries and within-five-second performance for report calculations.
- Used VIP-based automatic failover with a 1+3 topology and enabled backup execution from standby nodes.
- Completed migration of tens of terabytes of data, millions of PL/SQL lines, and more than 27 million objects with zero data loss and zero go-live errors.
- Added tuning parameters and delayed-standby protection to improve operational resilience.
Customer Value
- Supported group-level database governance by enabling a broader domestic replacement strategy across Oracle, PostgreSQL, MySQL, and related workloads.
- Raised the user’s operating, monitoring, and lifecycle-management capabilities through unified instance management.
- Reduced operational pressure for DBAs and developers by using delayed standby to protect against accidental operations.
- Enabled standby-based backup during 24x7 production without interfering with online services.