Write Filter Continuous Data Protection Snapshot
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems face challenges in reducing the Recovery Point Objective (RPO) without impacting performance, leading to significant data loss when a primary system goes offline, as they rely on infrequent snapshots and inefficient data transfer processes.
Innovation Solution
Implementing continuous data protection (CDP) through a write filter that intercepts and logs all input/output operations with sequence numbers, allowing for frequent data backup without performance degradation, and using a tree data structure to generate reference restoration points efficiently, reducing data loss by streamlining the snapshot process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If snapshots are performed more frequently to reduce RPO, then data loss is reduced, but system performance deteriorates
Solution Approach 1:
The patent segments the data protection process by implementing a write filter that operates independently at the driver level, separating snapshot creation from normal system operations. This allows frequent snapshots without impacting overall system performance by isolating the backup mechanism.
Solution Approach 2:
The write filter performs preliminary actions by intercepting and logging write operations before they are committed to the storage device. This pre-capturing of data changes enables rapid snapshot creation without requiring actual data transfer at snapshot time, thus maintaining system performance.
2Ease of manufacture
If traditional snapshot methods are used, then implementation is simple, but data transfer efficiency is low
Solution Approach 1:
The write filter performs preliminary logging of all write operations with sequence numbers before snapshots are needed. This pre-prepared log enables instant snapshot creation by simply pointing to the logged data, eliminating time-consuming data transfer while keeping the system simple to implement.
Solution Approach 2:
Instead of physically copying data for each snapshot, the system creates reference copies that point to the original logged data. This virtual copying mechanism maintains implementation simplicity while dramatically improving data transfer efficiency by avoiding redundant data movement.
3Productivity
If infrequent snapshots are used, then system performance is maintained, but data loss increases
Solution Approach 1:
The write filter continuously logs write operations in the background without interrupting system performance. This preliminary continuous recording ensures that even if snapshots are taken infrequently, no data is lost because all changes are already captured and sequenced in the log.
Solution Approach 2:
The write filter acts as an intermediary layer between the system and storage device, capturing all data changes without affecting system performance. This mediator continuously records data in the background, enabling reliable data protection with infrequent snapshot operations.
Data Source
AI summary
A reference snapshot of a storage is stored. Data changes that modify the storage are received. The data changes are captured by a write filter of the storage. The received data changes are logged. The data changes occurring after an instance time of the reference snapshot are applied to the reference snapshot to generate a first incremental snapshot corresponding to a first intermediate reference restoration point. The data changes occurring after an instance time of the first incremental snapshot are applied to the first incremental snapshot to generate a second incremental snapshot corresponding to a second intermediate reference restoration point.