Virtual Connection for Storage Mirroring Resynchronization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data processing systems that combine synchronous and asynchronous mirroring face challenges in maintaining data integrity and efficiency, particularly when links are suspended, leading to potential corruption or unusable data and the need for full resynchronization upon site failures.
Innovation Solution
Implementing a virtual connection between storage sites that allows for monitoring of data transfers without transmitting user data, enabling rapid re-establishment of mirroring relationships without full resynchronization, even in cascaded or fan-out configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If asynchronous mirroring is used to improve operating speed and reduce performance impact, then data can travel long distances, but site A lacks information about what data site B has sent to site C, requiring full resynchronization upon site B failure
Solution Approach 1:
The patent introduces an intermediary mechanism where site B (the intermediate site in the cascade) forwards data transfer information to site A. This intermediary role allows site A to obtain information about what data site B has sent to site C without requiring direct communication between A and C, thus resolving the information gap while maintaining the asynchronous mirroring architecture.
Solution Approach 2:
The patent implements a feedback mechanism where data transfer status information flows back from site C through site B to site A. This feedback loop enables site A to track what data has been replicated downstream, allowing it to perform selective resynchronization rather than full resynchronization when site B fails, thereby resolving the information loss problem.
2Reliability
If full resynchronization is performed upon site failure, then data integrity is restored, but downtime and cost increase significantly
Solution Approach 1:
The patent applies preliminary action by maintaining data transfer information records at site A before failures occur. By continuously tracking what data has been replicated through the cascade configuration, site A is prepared in advance with the information needed to perform selective resynchronization, eliminating the need for time-consuming full resynchronization when failures occur.
Solution Approach 2:
The patent changes the parameter of resynchronization scope from 'full' to 'selective' based on the information available. By using the tracked data transfer information to identify which specific data blocks need resynchronization rather than resynchronizing all data, the system reduces downtime while maintaining data integrity through targeted parameter adjustment.
3Length of stationary object
If cascaded mirroring configuration is used to extend replication distance, then data can be replicated across multiple sites, but site A cannot establish direct mirroring relationship with site C without full resynchronization
Solution Approach 1:
The patent makes site B serve multiple functions: it acts as both a data replication target from site A and a data source to site C, while simultaneously functioning as an information forwarder that relays data transfer status back to site A. This multi-functionality allows the cascaded configuration to maintain direct logical mirroring relationships without requiring full resynchronization.
Data Source
AI summary
In a storage system, at least three storage sites are arranged in a full mirroring relationship using a virtual connection between instances of a storage group on two of the sites whereby no user data is transmitted.