Virtual Machine Recovery Using Shared Data Chunks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data backup and recovery systems face inefficiencies when restoring multiple virtual machines, as they often require recreating and restoring equal amounts of data for each machine individually, leading to increased costs and resource utilization in on-demand computing environments.
Innovation Solution
A system that identifies and leverages similarities between virtual machines by designating a primary set of recovery data, replicating this data across similar machines, and only rehydrating differing data chunks for each virtual machine, thereby reducing the computational load and costs associated with data recovery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple virtual machines are recovered individually by recreating and restoring equal amounts of data for each machine, then each virtual machine can be restored to its original state, but the computational load and costs in on-demand computing environments increase significantly
Solution Approach 1:
The patent merges the recovery processes of multiple virtual machines by identifying and restoring common data chunks shared across VMs. Instead of treating each VM recovery independently, the system combines recovery operations for VMs with similar data profiles, restoring shared data once and reusing it across multiple VMs, thereby reducing total computational load and resource consumption.
Solution Approach 2:
The recovery system achieves multi-functionality by creating a universal recovery approach that can handle multiple virtual machines simultaneously. The system identifies universal data patterns and common data chunks across different VMs, enabling a single recovery process to serve multiple VMs rather than requiring separate dedicated recovery processes for each machine.
2Reliability
If multiple virtual machines are recovered individually by recreating and restoring equal amounts of data for each machine, then data integrity is maintained, but the costs in on-demand computing environments increase
Solution Approach 1:
The patent merges data transfer operations by identifying and restoring common data chunks shared across multiple virtual machines. Instead of transferring complete data sets for each VM independently, the system combines data transfer operations by recognizing overlapping data patterns, thereby reducing total data transfer volume while maintaining data integrity through verified restoration of shared data.
Solution Approach 2:
The system recovers and reuses common data chunks that can be shared across multiple virtual machines. By identifying data that can be discarded from individual VM recovery processes and recovered as shared resources, the system reduces redundant data transfer and storage requirements, lowering costs while preserving data integrity through controlled data reuse.
3Reliability
If recovery processes create multiple recovery virtual machines simultaneously for testing before production deployment, then system reliability is improved through testing, but on-demand computing resource costs increase
Solution Approach 1:
The patent merges recovery testing operations by conducting tests on shared data chunks and common recovery processes rather than independently testing each virtual machine recovery. This approach allows multiple VMs to be recovered and tested simultaneously using shared resources, improving resource utilization efficiency while maintaining recovery reliability through comprehensive testing of the unified recovery process.
Data Source
AI summary
Described is a system that recovers multiple virtual machines (or other computing resource) by leveraging similarities between recovery (or backup) data. The system may identify the recovery data for a particular virtual machine as a primary set of data using specialized calculations and/or heuristics. The system may leverage this primary set of data by replicating a data storage for the remaining number of virtual machines to be recovered. The system may then determine a remaining (or differing) set of data for each of the remaining virtual machines to be recovered. Accordingly, the system may only be required to rehydrate the differing set of data chunks and write those data chunks to the data storage associated with each recovery virtual machine. These recovery virtual machines may then be used to restore corresponding virtual machines within a production environment.