Granular Block Replication for Fast Container-Level Data Restore
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage systems require substantial time to restore data due to restoring entire data sets, even when only a portion is corrupted, leading to downtime during disaster recovery.
Innovation Solution
Implement a system and method for granular replication that allows selective restoration of corrupted data containers while maintaining the rest of the volume available for access, using a rollback PCPI and line-efficient data transfer.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If entire data set restoration is performed, then data recovery completeness is improved, but restoration time increases substantially
Solution Approach 1:
The patent divides the data set into discrete data containers (files, LUNs, or aggregates) that can be independently identified and restored. The system uses metadata structures to track individual container states, allowing selective restoration of only corrupted containers rather than the entire data set, thereby reducing restoration time while maintaining recovery completeness for affected data.
Solution Approach 2:
The system performs partial restoration by identifying and restoring only the specific data containers that are corrupted or lost, rather than restoring the entire data set. This partial action approach reduces unnecessary data transfer and processing time while ensuring complete recovery of the affected portions.
2Reliability
If entire volume is fenced for restoration, then data consistency is improved, but system availability deteriorates
Solution Approach 1:
The patent implements fencing at the data container level rather than volume level. Each corrupted data container is individually fenced to maintain consistency during restoration, while other containers within the same volume remain accessible. This granular fencing approach preserves data consistency for restored containers without preventing access to unaffected data, thereby maintaining system availability.
Solution Approach 2:
The system applies different states to different portions of the volume: fenced state for containers being restored and accessible state for containers remaining available. This local quality differentiation allows simultaneous data consistency maintenance and system availability, as clients can continue accessing non-corrupted containers while restoration proceeds on affected containers.
3Productivity
If granular restoration is implemented, then restoration efficiency is improved, but system complexity increases
Solution Approach 1:
The patent employs a universal metadata structure that can represent multiple data container types (files, LUNs, aggregates) using the same data model and management mechanisms. This universal approach enables granular restoration across different container types without requiring separate complex management systems for each type, thereby improving restoration efficiency while controlling system complexity through standardized interfaces.
Data Source
AI summary
A system and method for efficiently restoring one or more data containers is provided. A common persistent consistency point image (PCPI) is identified between a source and a destination storage systems prior to the destination storage system performing a rollback operation to the commonly identified PCPI. Differential data is then transmitted from the source storage system in a line efficient manner to the destination storage system.