Grid Server Deduplication Rate Adjustment for Backup Windows

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional deduplication-based data storage systems face availability issues due to reliance on a single compute server for deduplication, leading to increased end-to-end backup times as data volume grows, exceeding backup window limits and violating service level agreements.

Innovation Solution

A scalable server grid system dynamically adjusts deduplication and replication rates based on ingestion rates, allowing parallel processing of ingestion, deduplication, and replication, with failover models to maintain availability and reduce recovery point objective times.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If a single compute server is used for deduplication processing, then device complexity is reduced, but reliability deteriorates and backup time increases as data volume grows

Engineering Contradiction:
Improvesystem structureVSAvoiddata availability
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The system divides the deduplication processing function into multiple independent compute servers organized in a grid. Each server handles a portion of the deduplication workload independently, eliminating the single point of failure and allowing the system to scale with data volume while maintaining reliability and constant backup windows.

Inventive Principle:
Principle #1Segmentation

2Quantity of substance

If data volume increases and additional disk storage units are added, then storage capacity is improved, but backup time increases beyond service level agreement limits

Engineering Contradiction:
Improvestorage capacityVSAvoidbackup time
Core Design Contradiction:
Quantity of substanceVSLoss of time

Solution Approach 1:

The system segments the deduplication workload across multiple compute servers in a grid architecture. As storage capacity increases with additional disk units, the compute grid expands proportionally to maintain constant backup windows, ensuring service level agreements are met regardless of data volume growth.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system dynamically scales the compute grid capacity to match storage capacity growth. The number of compute servers is adjusted based on the amount of data to be processed, maintaining optimal performance and constant backup windows as the system evolves over time.

Inventive Principle:
Principle #15Dynamics

3Loss of time

If deduplication processing capacity is increased to handle larger data volumes, then backup time is reduced, but device complexity and cost increase

Engineering Contradiction:
Improvebackup windowVSAvoidsystem architecture
Core Design Contradiction:
Loss of timeVSDevice complexity

Solution Approach 1:

Each compute server in the grid is designed to be multi-functional, handling both data ingestion and deduplication processing. This universal design allows flexible resource allocation and scaling without requiring specialized hardware, reducing overall system complexity while maintaining high performance.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentEP3098714B1Dynamic and optimized management of grid system resources
Publication Date: 2023.02.15 EXAGRID SYST
  • EP3098714B1 patent drawingFigure 1
  • EP3098714B1 patent drawingFigure 2a
  • EP3098714B1 patent drawingFigure 2b

AI summary

A system, a method, and a computer program product for or managing grid server system resource(s) are disclosed. A processor of a grid server in a plurality of grid servers determines an ingestion rate of data received by the grid server for a backup. The ingestion rate corresponds to a speed of transferring of the received data into a memory of the grid server. The processor determines a deduplication and replication rate of the received data. Deduplication represents a deduplication of the received data by the grid server. Replication represents replication of the deduplicated received data to another server in the plurality of servers. The processor adjusts the deduplication and replication rate based on the ingestion rate.