Application-Aware Virtual Machine Backup via Data Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current backup methods for virtual machines (VMs) and applications are inefficient, leading to duplicated data storage costs, lack of flexibility in application-level restores, and inadequate support for granular recovery and application-aware backups, resulting in higher storage costs and reduced recovery options.
Innovation Solution
A system that combines VM image-level and application-level backups using a thin agent and a virtual environment protection agent (VEPA) to perform application-aware, agentless, and point-in-time backups, allowing for independent restoration of applications without restoring the entire VM, while utilizing a de-duplication engine to optimize storage by removing duplicate data blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If VM image-level backup is performed, then the entire VM can be restored, but application data is backed up twice (once in VM backup and once in application backup), increasing storage costs
Solution Approach 1:
The backup system segments data into two distinct categories: application data blocks and VM image blocks. Application data is extracted and stored separately from the VM image, allowing independent backup and restoration of application data without duplicating the entire VM image. This segmentation eliminates redundant storage while maintaining both VM-level and application-level restore capabilities.
2Reliability
If VM image-level backup is performed, then complete VM can be recovered, but granular application-level recovery is not supported
Solution Approach 1:
The system segments the backup into application data blocks and VM image blocks, enabling granular recovery of individual applications or specific data blocks independent of the VM image. This allows users to restore only the necessary application data without recovering the entire VM, providing flexible and adaptable recovery options.
Solution Approach 2:
The system adds a new dimension to the backup architecture by creating an independent application data layer separate from the VM image layer. This dimensional separation enables recovery operations at multiple levels: full VM restoration, selective application restoration, or even individual data block restoration, significantly enhancing recovery flexibility.
3Adaptability or versatility
If application-level backup is performed separately from VM backup, then application data can be restored independently, but data is stored twice leading to higher costs
Solution Approach 1:
The system merges the application-level backup and VM image backup into a unified backup architecture. Application data blocks are extracted from the VM image and stored in an application data store, while the VM image is stored separately. This combination allows independent application restoration without duplicating storage, as the system intelligently references the shared storage infrastructure to avoid redundant data copies.
Data Source
AI summary
A method relates to backing up a VM. The method can include placing an application being hosted by a VM in a backup state and capturing a snapshot of the VM to generate a VM image of the VM. The method can further comprise resuming the application, extracting application blocks corresponding to the application from the mounted VM image, and storing the application blocks and the VM image.