Backup Replication via Differential Data Segmentation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Replicating full backups in deduplicating storage systems is inefficient due to substantial overhead and bandwidth requirements, even when using deduplication techniques, as it involves transmitting identifying information for each data chunk of a full backup.

Innovation Solution

The system calculates the differential between the new backup file and the most recent file previously replicated, segments the difference data, and transmits segment identifying information, allowing the replica system to recreate the backup file by referencing unchanged regions from the previous backup and transmitting only new data segments.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If full backup replication is performed in a deduplicating storage system by transmitting identifying information for each data chunk, then data safety is improved through replication, but transmission overhead and bandwidth requirements increase substantially

Engineering Contradiction:

Improvedata safetyVSAvoidtransmission overhead

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The backup file is divided into multiple data segments, and only the identifying information (fingerprints) of these segments is transmitted to the replica system. The replica system then reconstructs the full backup by retrieving segments from the primary system using these identifiers, significantly reducing transmission overhead while maintaining data safety

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Instead of transmitting actual backup data or all segment identifiers, the system transmits only fingerprint copies (identifying information) of data segments. These fingerprints serve as references that enable the replica system to reconstruct the full backup without receiving the actual data chunks, thereby reducing bandwidth consumption

Inventive Principle:

Principle #26Copying

2Reliability

If full backup replication is performed by transmitting identifying information for each data chunk, then complete data replication is achieved, but replication time increases due to substantial overhead

Engineering Contradiction:

Improvedata replication completenessVSAvoidreplication time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent extracts only the essential identifying information (fingerprints) from the complete backup data and transmits only these extracted identifiers to the replica system. This extraction approach maintains replication completeness while dramatically reducing transmission time and overhead

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The backup file is pre-segmented into data segments with assigned fingerprints before transmission. This preliminary organization allows the replica system to efficiently retrieve only necessary segments using the pre-computed identifiers, reducing replication time compared to processing complete backups or performing segmentation after transmission

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS9465695B2Efficient backup replication

Publication Date: 2016.10.11 EMC IP HLDG CO LLC

AI summary

A system for backup replication comprises a processor and a memory. The processor is configured to determine data present in a most recent backup not present in a previous backup; transmit an extent specification; and transmit data segment fingerprints of the one or more data segments. The memory is coupled to the processor and is configured to provide the processor with instructions.