Efficient Backup and Disaster Recovery

Traditional database systems use file or block storage, which means that backup and restoration speed depends on the data volume. If there is a large amount of data, quick backup and restoration cannot be achieved, and data reliability is impacted.

GeminiDB uses the append-only storage in the DFV storage pool and optimizes the distributed parallel algorithm based on the database logic, improving the backup and restoration performance.

Data Function Virtualization (DFV) is a high-performance and high-reliability distributed storage system that is vertically integrated with databases. Storage clusters are deployed in pools to improve storage utilization.

• Powerful data snapshot processing capability: Data is stored in append-only mode in multiple copies at multiple time points. Snapshots are generated within seconds, and massive amounts of snapshots can be processed.
• Parallel high-speed backup and restoration: Backup restoration logic is pushed down to storage nodes. Data is accessed locally and directly interacted with storage systems, improving database performance and concurrency.
• Quick instance recovery
• Full functions are available in minutes leveraging the asynchronous data synchronization and on-demand real-time data loading mechanism.