Storage Apparatus Dump Data Deduplication for Failure Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The large size of dump data in storage apparatuses poses a challenge for efficient management, requiring significant storage capacity and complicating immediate data acquisition during failures, which is critical for maintaining business continuity.

Innovation Solution

A storage apparatus with a processor and memory that executes storage processing to store dump data and applies deduplication techniques to reduce data redundancy, thereby managing dump data more efficiently.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If dump data is stored for many generations to enable failure analysis, then the ability to analyze failures is improved, but the storage capacity requirement increases significantly

Engineering Contradiction:
Improvefailure analysis capabilityVSAvoidstorage capacity
Core Design Contradiction:
ReliabilityVSVolume of stationary object

Solution Approach 1:

The patent creates a copy of the dump data in a non-volatile memory device (such as a flash memory or SSD) separate from the main memory. This copy allows the system to retain historical dump data for failure analysis without requiring the main memory to continuously hold all data, thereby reducing the immediate storage capacity burden while maintaining the ability to analyze past failures.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent divides the dump data management into different storage layers: volatile memory for active processing and non-volatile memory for historical storage. This segmentation allows the system to manage dump data across multiple generations by storing older dumps in the non-volatile memory, reducing the space requirement in any single storage location while maintaining comprehensive analysis capability.

Inventive Principle:
Principle #1Segmentation

2Loss of information

If maintenance operation is performed to acquire dump data, then the data acquisition is possible, but the business recovery time increases

Engineering Contradiction:
Improvedump data acquisitionVSAvoidbusiness recovery time
Core Design Contradiction:
Loss of informationVSLoss of time

Solution Approach 1:

The patent implements automatic dump data acquisition where the storage apparatus itself captures and stores dump data in the non-volatile memory device without requiring manual intervention from maintenance personnel. The system automatically detects failures, acquires relevant memory data, and stores it for later analysis, eliminating the need for maintenance operations to manually access and copy dump data, thereby reducing business recovery time.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent performs preliminary actions by continuously storing dump data in the non-volatile memory device as it occurs, rather than waiting for a maintenance operation to initiate data acquisition. This ensures that dump data is already available and ready for analysis immediately when a failure is detected, eliminating delays associated with manual data acquisition procedures.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS11294751B2Storage apparatus, dump data management method, and dump data management program
Publication Date: 2022.04.05 HITACHI VANTARA LTD
  • US11294751B2 patent drawing
  • US11294751B2 patent drawing
  • US11294751B2 patent drawing

AI summary

Dump data of a memory is allowed to be easily and appropriately managed. In a storage apparatus including a CPU, a DRAM, and a drive, the CPU executes storage processing to store data of the DRAM to be stored into the drive as dump data upon a failure occurrence in the storage apparatus. The CPU deduplicates the dump data in the storage processing or after storing the dump data into the drive. The CPU may execute deduplication processing between the dump data and other dump data stored in the drive or may execute deduplication processing within the data to be stored.