Synthetic Backup Metadata for Efficient Incremental Recovery

Solution Overview

Problem

Traditional backup systems face performance issues and resource consumption when dealing with large numbers of incremental backups, as they require significant computing resources and storage space for synthetic backups, especially when attempting granular point-in-time recoveries.

Innovation Solution

The method involves creating synthetic views of incremental backups by combining block metadata from incremental backups with previous backups, generating new synthesized block metadata that identifies the location of the most current version of each block without copying data blocks, allowing for efficient recovery without the need for extensive synthetic backups.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional backup systems create synthetic backups to manage incremental backups, then recovery performance improves, but computing resources and storage space consumption increase significantly

Engineering Contradiction:

Improverecovery performanceVSAvoidcomputing resources consumption

Core Design Contradiction:

ProductivityVSUse of energy by stationary object

Solution Approach 1:

The patent creates synthetic views of backup data by combining metadata from incremental backups with metadata from previous backups, rather than copying actual data blocks. This allows the system to present a complete view of data at any point in time without physically replicating the data itself, thereby improving recovery performance while avoiding the storage and computing resource consumption associated with traditional synthetic backup creation.

Inventive Principle:

Principle #26Copying

2Productivity

If traditional backup systems create synthetic backups to manage incremental backups, then recovery performance improves, but storage space consumption increases significantly

Engineering Contradiction:

Improverecovery performanceVSAvoidstorage space consumption

Core Design Contradiction:

ProductivityVSVolume of stationary object

Solution Approach 1:

The patent creates synthetic views of backup data by combining metadata from incremental backups with metadata from previous backups, rather than copying actual data blocks. This allows the system to present a complete view of data at any point in time without physically replicating the data itself, thereby improving recovery performance while avoiding the storage and computing resource consumption associated with traditional synthetic backup creation.

Inventive Principle:

Principle #26Copying

3Reliability

If backup systems preserve indefinitely long series of incremental backups, then data protection coverage improves, but recovery time and computing resources increase

Engineering Contradiction:

Improvedata protection coverageVSAvoidrecovery time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent pre-computes and stores synthetic metadata for incremental backups, combining block metadata from incremental backups with synthesized block metadata from previous backups. This preliminary action allows the system to quickly reconstruct complete data views at any point in time during recovery operations, significantly reducing recovery time even when dealing with indefinitely long series of incremental backups while maintaining comprehensive data protection coverage.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS9430332B1Systems and methods for enabling efficient access to incremental backups

Publication Date: 2016.08.30 COHESITY INC

AI summary

A computer-implemented method for enabling efficient access to incremental backups may include selecting an incremental backup of a volume of data, the incremental backup of the volume of data storing blocks from the volume of data that have changed since a previous backup of the volume of data, identifying block metadata for the incremental backup that describes a location of each block within the blocks in the incremental backup, detecting synthesized block metadata for the previous backup that describes a location of each block in the volume according to a synthesized view of the volume of data from a full backup of the volume up to the previous backup, and combining the block metadata for the incremental backup with the synthesized block metadata for the previous backup to generate new synthesized block metadata for the incremental backup. Various other methods, systems, and computer-readable media are also disclosed.