Storage Subsystem Controller Segmentation for Fault Tolerance

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional storage subsystems do not effectively manage faults in controllers, leading to performance deterioration when a fault occurs, as they often require replacing entire controllers, causing intensive access on other controllers and potential response delays.

Innovation Solution

A storage subsystem with multiple storage apparatuses connected via a network, where each apparatus is divided into systems with controllers, interface modules, and subordinate interfaces, allowing for the selection of normal resources to continue data processing even if abnormalities occur, and requesting other systems to handle affected tasks via a data transfer path.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If the entire controller is replaced when a fault occurs in part of it, then the faulty element is removed, but the normal elements are also discarded and the other controller becomes overloaded

Engineering Contradiction:
Improvefault handlingVSAvoidprocessing capacity
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The controller is divided into functional elements (interface module, subordinate interface, controller logic) that can be independently assessed for faults. When a fault occurs, only the specific faulty element is identified and isolated, while other normal elements continue to function, preventing unnecessary replacement of the entire controller and maintaining processing capacity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system applies different quality assessments to different parts of the controller based on their operational status. Normal elements continue to process data while faulty elements are isolated. This local differentiation allows the system to maintain overall functionality by preserving normal segments while addressing only the specific problematic areas.

Inventive Principle:
Principle #3Local quality

2Reliability

If the host computer switches to access storage devices via the other controller, then access continuity is maintained, but response performance deteriorates due to increased internal processing time

Engineering Contradiction:
Improveaccess continuityVSAvoidresponse performance
Core Design Contradiction:
ReliabilityVSSpeed

Solution Approach 1:

The system introduces an intermediary mechanism that monitors controller element status and dynamically routes commands. When a fault is detected, the intermediary redirects host computer access to alternative normal elements within the same controller or to the other controller only when necessary, minimizing the performance penalty by keeping communication paths as direct as possible.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Device complexity

If the controller integrates all elements, then the structure is simplified, but a fault in one element blocks the entire controller requiring full replacement

Engineering Contradiction:
Improvecontroller structureVSAvoidfault tolerance
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The controller architecture is segmented into distinct functional modules (interface module for host communication, subordinate interface for storage device communication, controller logic for command processing). This segmentation allows the system to maintain simplified overall structure while enabling selective isolation of faulty modules, thereby achieving both structural simplicity and fault tolerance.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS8756381B2Storage subsystem and load distribution method for executing data processing using normal resources even if an abnormality occurs in part of the data processing resources that intermediate data processing between a host computer and a storage device
Publication Date: 2014.06.17 HITACHI VANTARA LTD
  • US8756381B2 patent drawing
  • US8756381B2 patent drawing
  • US8756381B2 patent drawing

AI summary

A storage subsystem coupled to a host computer is described. The storage subsystem includes storage devices and first and second storage apparatuses that control data transfer between the host computer and the storage devices. The first storage apparatus includes a first controller coupled to the host computer via a first host communication control unit and to the storage devices via a first storage device communication control unit. The second storage apparatus includes a second controller coupled to the host computer via a second host communication control unit and to the storage devices via a second storage device communication control unit. At least one of the controllers monitors a status of the first host communication control unit and the storage device communication control units, and, if the status of the first storage device communication unit indicates failure, switch communication paths for transferring data from the host computer to the storage devices.