Storage Network Failure Recovery Using Snapdifference Files
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage systems face limitations in providing fast and space-efficient copy operations for disk volumes, especially in scalable, distributed, and table-driven virtual storage systems, and require improved methods for failure recovery in storage devices.
Innovation Solution
The implementation of a method in a storage network that detects failures in source volumes, terminates communication, refreshes the volume, copies a backup data set, activates a snapdifference file, and records I/O operations, allowing for efficient recovery and data management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a complete copy of the disk volume is made, then data recovery capability is improved, but storage space consumption increases
Solution Approach 1:
The patent segments the backup storage into two distinct components: a base image containing the original disk volume data, and a snapdifference file containing only the changes (differences) from the base image. This segmentation allows the system to store complete backup data while using minimal additional space for updates, as the snapdifference file only records modified blocks rather than duplicating entire volume data.
Solution Approach 2:
The patent implements a copying mechanism where instead of creating complete duplicate copies of the disk volume for each backup point, the system creates a base image copy and then uses differential copying for subsequent backups. The snapdifference file stores only the differences between the base image and current state, dramatically reducing storage space requirements while maintaining full data recovery capability.
2Reliability
If a disk copy operation is performed, then data redundancy is improved, but I/O operation performance deteriorates
Solution Approach 1:
The patent performs preliminary actions by creating a base image of the disk volume in advance, before any differential backups are needed. This base image serves as a static reference that does not require frequent updates, allowing the system to establish data redundancy upfront while minimizing subsequent I/O operations. The base image can be created during low-utilization periods without impacting ongoing application performance.
Solution Approach 2:
The patent applies partial action by only copying and storing the portions of data that have changed since the base image was created. Instead of performing complete disk copy operations for each backup, the system identifies and copies only the modified blocks, significantly reducing I/O overhead and performance impact while maintaining adequate data redundancy for recovery purposes.
3Reliability
If traditional disk copy methods are used, then data backup capability is improved, but scalability in distributed systems deteriorates
Solution Approach 1:
The patent creates a universal backup architecture where the base image and snapdifference file mechanism can serve multiple functions across distributed systems. The same infrastructure supports data backup, incremental updates, point-in-time recovery, and space-efficient archiving. This multi-functional approach enhances scalability by providing a standardized solution that can be deployed across multiple nodes and storage systems without requiring different mechanisms for different purposes.
Data Source
AI summary
Exemplary storage network architectures, data architectures, and methods for creating and using snapdifference files in storage networks are described. One exemplary method may be implemented in a processor in a storage network. The method comprises detecting a failure in a source volume, and in response to the failure: terminating communication with one or more applications that generate I/O requests to the source volume; refreshing the source volume; copying a backup data set to the source volume, and while the backup data set is being copied: activating a new snapdifference file; restarting communication with one or more applications that generate I/O requests to the source volume; and recording I/O operations to the source volume in the snapdifference file.