Rapid Failback Using Write Logs and Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional failback operations in disaster recovery systems are slow due to the need for synchronization of primary and secondary data storage, preventing access to all data until synchronization is complete.
Innovation Solution
A method and system that utilize a write log and a map to allow read requests during failback, directing them to the latest version of data in either the primary or secondary data storage, enabling access to data without waiting for synchronization to finish.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data synchronization between primary and secondary data storage is completed before allowing read requests during failback, then data consistency is ensured, but failback time is prolonged
Solution Approach 1:
The system performs preliminary actions by maintaining a write log on the primary computer and a map on the secondary computer that tracks data locations before failback occurs. This allows the system to quickly determine where data resides during failback without needing to complete full synchronization first, thus reducing failback time while maintaining data consistency through the pre-established tracking mechanisms
Solution Approach 2:
The patent introduces intermediary structures (write log and map) that mediate between primary and secondary data storage. These intermediaries track data locations and versions, enabling the system to serve read requests correctly during failback by consulting the map and write log rather than waiting for complete synchronization, thus resolving the contradiction between speed and consistency
2Reliability
If applications wait for storage synchronization to complete before resuming operations, then data integrity is maintained, but productivity is reduced
Solution Approach 1:
The system prepares in advance by maintaining write logs and maps that track all data locations and versions. When failback occurs, applications can immediately resume operations by querying these pre-established structures to determine data locations, rather than waiting for synchronization to complete. This preliminary preparation enables both rapid application resume and data integrity maintenance
Solution Approach 2:
The map and write log provide continuous feedback about data locations and versions. During failback, applications can query this feedback mechanism to determine where to read data from, ensuring they access the latest versions without waiting for full synchronization. This feedback loop maintains data integrity while enabling immediate application operation
Data Source
AI summary
A method, system and computer-readable medium for providing rapid failback of a computer system is described. The method, which operates during failback of a secondary computer to a primary computer, accesses a map to determine a location of a latest version of data corresponding to a read request, where the location may be within either a primary data storage or a secondary data storage. The system comprises a primary computer coupled to a primary data storage and a secondary computer coupled to a secondary data storage. The primary computer maintains a write log and the secondary computer maintains a map. The computer-readable medium contains instructions, which, when executed by a processor, performs the steps embodied by the method.