Incremental Backup Reference Map for Fast Restoration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Incremental backups in deduplication systems complicate restoration operations due to the lack of complete data items, requiring processing of multiple incremental backups, which is inefficient compared to full backups.

Innovation Solution

The method involves initiating an incremental backup, sending new or changed data items to a deduplication engine, generating fingerprints, and creating a reference map that includes references to unchanged data items, allowing for efficient data retrieval and restoration by referencing a single extents catalog and reference map, regardless of the number of incremental backups since the last full backup.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If incremental backups are performed to reduce storage space and backup time, then storage utilization decreases and backup processing time reduces, but restoration operations become more complex and less efficient

Engineering Contradiction:
Improvestorage spaceVSAvoidrestoration complexity
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The system performs preliminary actions by creating a complete extents catalog during the incremental backup process that maps all data items (both changed and unchanged) to their locations. This pre-established catalog enables efficient restoration without requiring complex processing of multiple incremental backup chains, thus resolving the contradiction between reduced storage and increased restoration complexity

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary mechanism - the extents catalog and reference map system - that mediates between the incremental backup storage structure and the restoration process. This intermediary contains complete references to all data items across incremental backups, allowing simple restoration operations that function as if restoring from a full backup, thereby reducing restoration complexity while maintaining incremental backup storage efficiency

Inventive Principle:
Principle #24Intermediary (Mediator)

2Loss of time

If incremental backups are performed to reduce backup processing time, then backup time decreases, but restoration efficiency decreases compared to full backups

Engineering Contradiction:
Improvebackup processing timeVSAvoidrestoration efficiency
Core Design Contradiction:
Loss of timeVSProductivity

Solution Approach 1:

During the incremental backup process, the system performs preliminary actions by building and updating the extents catalog and reference maps that contain complete data item information. This preliminary work during backup enables fast restoration operations later, decoupling the time benefits of incremental backups from restoration efficiency, thus resolving the contradiction between reduced backup time and maintained restoration efficiency

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent creates a copied reference structure (extents catalog and reference map) that replicates the comprehensive data item information found in full backups. This copied reference enables restoration operations to proceed with the same efficiency as full backup restoration, while the actual backup process benefits from incremental time savings, thereby resolving the contradiction between backup time reduction and restoration efficiency

Inventive Principle:
Principle #26Copying

Data Source

PatentUS8825972B1Method and system of producing a full backup image using an incremental backup method
Publication Date: 2014.09.02 COHESITY INC
  • US8825972B1 patent drawing
  • US8825972B1 patent drawing
  • US8825972B1 patent drawing

AI summary

A system, method, and medium for performing incremental backups of a data set within a deduplication storage system. An incremental backup of a data set is initiated, and the changed data items of the data set are deduplicated. An extents catalog is created to fully describe the data set. References associated with the unchanged data items of the data set are retrieved from the extents catalog and then sent to a deduplication engine. The deduplication engine uses the references to retrieve fingerprints associated with the unchanged data items from a previous reference map. Then, the deduplication engine creates a new reference map for the incremental backup and stores the newly generated and retrieved fingerprints in the new reference map.