Zero Data Bypass in Thin Logical Unit Replication
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 that determines if an I/O data block from a journal is generated by a replication initialization process, and if it contains zero data, bypasses writing it to the volume, optimizing data replication and recovery by efficiently managing journal entries and data streams to enable rollback to any specified point in time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data replication is performed by writing all I/O data blocks from journal to volume, then data consistency is ensured, but processing time and overhead increase
Solution Approach 1:
The patent changes the parameter being monitored from raw data content to data signature (hash value). By computing and comparing signatures of I/O data blocks instead of comparing actual data content, the system maintains data consistency verification while significantly reducing processing time and overhead, especially for thin logical units where many blocks may be zero or unchanged.
2Reliability
If conventional data protection systems perform backup operations, then data protection is achieved, but system shutdown is required during backup
Solution Approach 1:
The patent performs preliminary data replication by maintaining a journal of I/O operations and applying them to the remote logical unit in advance. This allows the backup site to be continuously updated without requiring system shutdown, enabling continuous data protection while maintaining system availability at both primary and backup sites.
3Quantity of substance
If data is backed up on a daily basis, then storage costs are reduced, but several hours of data loss occur during disaster recovery
Solution Approach 1:
The patent implements continuous data replication by continuously capturing I/O operations in a journal and continuously applying them to the remote logical unit. This continuous action ensures that the backup site maintains an up-to-date copy of data, enabling recovery to any recent point in time rather than only to daily backup points, thus reducing both storage requirements and recovery time.
4Reliability
If the data recovery process is performed using conventional systems, then data restoration is achieved, but the process takes a long time
Solution Approach 1:
The patent performs preliminary replication of all data changes to the remote site before disasters occur. By maintaining a continuously updated remote logical unit through journal playback, the system prepares recovery data in advance, allowing rapid restoration by simply switching to the pre-prepared remote copy rather than performing lengthy recovery operations after a disaster.
Data Source
AI summary
In one aspect, a method to replicate data includes determining if an I/O data block from a journal is generated by a replication initialization process; if the I/O data block from a journal is generated by the replication initialization process, determining if the I/O data block from the journal comprises zero data; if the I/O data block from the journal comprises the zero data, reading an I/O data block in a volume; and determining if the I/O data block from the volume comprises zero data. The method also includes, if the I/O data block from the volume comprises zero data, bypassing writing the I/O data block from the journal to the I/O data block in the volume.