Dynamic Memory Redundancy Cancellation for Multi-CPU Systems

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current computer systems face challenges in maintaining memory redundancy during CPU failures, leading to reduced available memory size and potential system startup failures, especially in multi-CPU configurations with memory mirroring, which does not compensate for memory reduction effectively.

Innovation Solution

An information processing apparatus and method that includes a normality checker, failure detector, isolation unit, redundancy determiner, and redundancy canceller to assess and cancel memory redundancy configurations, ensuring available memory size approximates the duplexed state before CPU fallback, thereby maintaining system performance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If memory mirroring is implemented to ensure redundancy, then system reliability is improved, but available memory size is reduced

Engineering Contradiction:
Improvesystem reliabilityVSAvoidavailable memory size
Core Design Contradiction:
ReliabilityVSQuantity of substance

Solution Approach 1:

The patent implements dynamic memory configuration that automatically adjusts between mirroring and non-mirroring modes based on system state. When CPU failure is detected, the system dynamically cancels mirroring for remaining CPUs, transforming the static memory configuration into an adaptive system that responds to failure conditions.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the memory mirroring parameter from a fixed state to a variable state. The memory controller modifies the mirroring configuration parameter based on CPU operational status, switching between mirrored and non-mirrored modes to optimize available memory while maintaining reliability where needed.

Inventive Principle:
Principle #35Parameter changes

2Stability of the object's composition

If CPU isolation is performed upon failure detection, then system stability is improved, but available memory size is reduced

Engineering Contradiction:
Improvesystem stabilityVSAvoidavailable memory size
Core Design Contradiction:
Stability of the object's compositionVSQuantity of substance

Solution Approach 1:

The patent creates a dynamic response to CPU failure that goes beyond static isolation. The system automatically detects CPU failure, isolates the failed CPU, and then dynamically reconfigures memory settings for remaining CPUs to cancel mirroring, thereby recovering memory capacity while maintaining system stability through the isolation mechanism.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system implements a feedback loop where CPU failure detection triggers automatic isolation and memory reconfiguration. The memory controller receives feedback about CPU status and automatically adjusts memory configuration in response, creating a closed-loop control system that maintains stability while optimizing resource utilization.

Inventive Principle:
Principle #23Feedback

3Reliability

If memory redundancy is maintained after CPU failure, then reliability is preserved, but memory management table bloats and application startup may fail

Engineering Contradiction:
ImprovereliabilityVSAvoidmemory management table complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent changes the memory configuration parameter from mirrored to non-mirrored mode based on CPU failure status. This parameter change reduces the complexity of the memory management table by eliminating redundant entries, thereby preventing table bloat and avoiding application startup failures while maintaining appropriate reliability through selective redundancy cancellation.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP2778934B1Information processing device, information processing method, information processing program, and recording medium in which program is recorded
Publication Date: 2021.05.12 FUJITSU LTD
  • EP2778934B1 patent drawingFigure 1
  • EP2778934B1 patent drawingFigure 2
  • EP2778934B1 patent drawingFigure 3

AI summary

An information processing apparatus includes: processors (3) each having a memories (5) and a memory controller that controls the memories (5); a normality checker (131) that checks whether the processors (3) operate normally when started; a failure detector (133) that finds any failed processor (3) from a result of the check; a fallback unit (135) that falls back a failed processor (3) if any; a redundancy determiner (137) that determines whether the memories (5) are used in a redundancy configuration; a redundancy cancellation determiner (139) that determines, when the memories (5) are determined to be used in the redundant configuration, whether the redundancy configuration of the memories (5) is to be cancelled; and a redundancy canceller (141) that cancels, when the redundancy configuration of the memories (5) is to be cancelled, the redundancy configuration of the memories (5) in at least one processor (3) operating normally.