Virtual Machine Disk Replication via Journal-Based Cloud Snapshots

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional data protection systems face challenges in efficiently replicating and snapshotting virtual machine disks across different sites, especially in cloud environments, where ensuring data consistency and rapid recovery in case of disasters is critical.

Innovation Solution

The method involves storing virtual machine disks in logical units on a data store, replicating them to a cloud array storage device, storing changes in a journal, and periodically rolling back to a consistent point-in-time using the journal to take snapshots and send them to a cloud storage network.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If data replication is performed continuously to maintain real-time data consistency between production and backup sites, then data reliability is improved, but network bandwidth consumption and system performance overhead increase

Engineering Contradiction:

Improvedata consistencyVSAvoidnetwork bandwidth consumption

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The system performs preliminary actions by maintaining a write journal locally before replication occurs. Change blocks are identified and prepared for replication in advance, allowing the system to batch and optimize network transmissions rather than continuously streaming data, thus reducing overall network bandwidth consumption while maintaining data consistency.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

Instead of replicating entire data sets continuously, the system creates selective copies of only the changed data blocks identified through journal analysis. This targeted copying approach maintains data consistency where needed while significantly reducing network bandwidth consumption by transmitting only necessary data portions.

Inventive Principle:

Principle #26Copying

2Productivity

If snapshots are taken frequently to ensure rapid disaster recovery, then recovery speed is improved, but storage overhead and system performance impact increase

Engineering Contradiction:

Improverecovery speedVSAvoidstorage overhead

Core Design Contradiction:

ProductivityVSQuantity of substance

Solution Approach 1:

The snapshot functionality is segmented into incremental updates rather than full system snapshots. By dividing the snapshot process into discrete change blocks tracked through the write journal, the system enables rapid recovery through accumulated incremental data while minimizing storage overhead compared to traditional full snapshots.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The write journal continuously tracks changes between snapshots, maintaining a continuous record of data modifications. This continuous tracking eliminates gaps in recovery capability while avoiding the need for frequent full snapshots, thus improving recovery speed without proportionally increasing storage overhead.

Inventive Principle:

Principle #20Continuity of useful action

3Reliability

If journaling is enabled to track all data changes for point-in-time recovery, then data protection capability is improved, but write performance and storage overhead increase

Engineering Contradiction:

Improvepoint-in-time recovery capabilityVSAvoidwrite performance

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The journal implementation uses copy-on-write technology where only changed data blocks are copied to the journal rather than duplicating entire data sets. This approach maintains comprehensive point-in-time recovery capability while minimizing the performance impact on write operations and reducing storage overhead by journaling only necessary change blocks.

Inventive Principle:

Principle #26Copying

4Reliability

If remote replication is performed across geographically distributed sites, then disaster recovery robustness is improved, but network latency and data lag increase

Engineering Contradiction:

Improvedisaster recovery robustnessVSAvoiddata lag

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

Change blocks are identified and prepared locally in the write journal before replication to remote sites. This preliminary local processing eliminates the need for real-time coordination during replication, allowing asynchronous transmission of prepared change blocks to remote sites, thus reducing data lag while maintaining disaster recovery robustness.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The replication system uses periodic batch transmissions of accumulated change blocks rather than continuous real-time replication. This periodic approach reduces network latency impact by aggregating changes and transmitting them in optimized batches, maintaining data consistency while reducing the perceived data lag across geographically distributed sites.

Inventive Principle:

Principle #19Periodic action

Data Source

PatentUS10067837B1Continuous data protection with cloud resources

Publication Date: 2018.09.04 EMC IP HLDG CO LLC

AI summary

In one aspect, a method includes storing a virtual machine disk in a first logical unit stored on a data store; replicating the first virtual machine disk used by a first virtual machine at a second virtual machine disk stored on a second logical unit on a cloud array storage device; storing changes to the first virtual machine disk on a journal stored on the data store; periodically rolling each virtual machine disk of virtual machines on the second logical unit to a same-point-in-time using the journal and taking a snapshot of the second logical unit at the same point-in-time and sending the snapshots to a cloud storage network.