Asynchronous Data Replication via Metadata Delta Marking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data protection systems face challenges such as system shutdown during backups, limited recovery points, and lengthy data recovery processes, which hinder continuous data protection and increase downtime in the event of disasters.
Innovation Solution
A method and apparatus that change a production site from one node to another, record a point in time of failover, and replicate data by determining the latest common point in time for metadata changes, adding metadata changes to delta marking streams on both nodes, and managing these changes to enable efficient data recovery and minimize downtime.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional tape backup systems are used, then data can be stored on periodic basis, but system shutdown is required during backup and data availability is reduced
Solution Approach 1:
The patent implements continuous data protection through asynchronous replication, where data changes are continuously copied to a secondary storage system without interrupting production operations. This eliminates the need for system shutdown during backup while maintaining data protection, resolving the contradiction between reliability and productivity.
Solution Approach 2:
The patent introduces a replication system as an intermediary between the production storage system and backup storage system. This intermediary continuously synchronizes data changes, enabling backup operations to occur without blocking production access, thus maintaining system availability while ensuring data protection.
2Reliability
If daily backup is performed, then data is protected at discrete time points, but recovery time window is limited and data loss occurs
Solution Approach 1:
The patent transitions from discrete daily backups to continuous replication, where data is continuously synchronized between primary and secondary storage systems. This continuous operation enables recovery to any point in time within the recent history, eliminating the fixed recovery time window limitation of daily backups while maintaining data protection.
Solution Approach 2:
The patent performs preliminary continuous replication of data changes before any disaster occurs. By maintaining an ongoing copy of all data changes in the recent history, the system prepares recovery data in advance, enabling immediate recovery to any specified point in time without waiting for the next scheduled backup.
3Reliability
If conventional backup systems are used, then data can be recovered, but recovery process takes a long time
Solution Approach 1:
The patent performs preliminary continuous replication of all data changes to the secondary storage system before any disaster occurs. This pre-positioned replicated data can be immediately restored without waiting for lengthy backup processes, significantly reducing recovery time while maintaining data protection.
Solution Approach 2:
The patent creates a continuous copy of data changes through asynchronous replication, maintaining a synchronized replica of the production storage system. This copy can be rapidly restored in case of disaster, reducing recovery time compared to traditional backup systems that require re-transmission of entire data sets.
4Productivity
If signature comparison is used for data synchronization, then efficiency is improved, but compatibility issues arise when signature comparison is not available
Solution Approach 1:
The patent changes the synchronization parameter from data signature comparison to metadata change tracking. This parameter change enables the system to operate efficiently by comparing only metadata changes rather than entire data blocks, while maintaining compatibility with systems where signature comparison is not available, thus resolving the contradiction between productivity and adaptability.
Data Source
AI summary
In one aspect, a method, includes changing a production site from a first node to a second node, recording a point in time, Pi, of the production site fail over from the first node to the second node and replicating data from the second node to a third node. The replicating includes determining a latest common point in time, Px, that metadata changes were received between the second node and the third node, adding metadata changes between the time, Px, and a latest point in time that metadata was received at the third node to a delta marking stream on the third node, determining an earliest time, PL, between the time, Pi, and the time, Px, and adding metadata changes from the time, PL, to the latest point in time that metadata was received at the second node to a delta marking stream on the second node.