Multi-Core Resource Isolation for Single Disk Failure Domains

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In multi-core ARM processor systems with multiple hard disks, the failure of the operating system results in a coarse failure recovery granularity, leading to increased system pressure due to exponential growth in data restoration and network transmission, as the failure domain spans multiple hard disks, compromising reliability and resource utilization.

Innovation Solution

A method and system for resource isolation in a multi-core architecture, where a processor with multiple cores, including a start-time primary core and secondary cores, uses a shared memory segment to store running information and run-time entry addresses, allowing each core to independently manage its kernel images and I/O operations, thereby isolating failures to a single hard disk and ensuring independent operation of operating systems on different cores.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If an operating system runs on a multi-core processor with multiple hard disks, then resource utilization is improved, but failure recovery granularity becomes coarse and system pressure increases

Engineering Contradiction:
Improveresource utilizationVSAvoidfailure recovery granularity
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent divides the multi-core processor into multiple isolated domains, where each core runs its own operating system instance with dedicated memory space and I/O resources. This segmentation ensures that failures are isolated to individual cores rather than affecting the entire system, thereby improving failure recovery granularity while maintaining high resource utilization through parallel operation of multiple OS instances.

Inventive Principle:
Principle #1Segmentation

2Quantity of substance

If multiple hard disks are used in a multi-core system, then storage capacity is improved, but failure domain expands across multiple disks

Engineering Contradiction:
Improvestorage capacityVSAvoidfailure domain
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent assigns specific hard disks to specific processor cores through isolated I/O resource allocation. Each core's operating system can only access its designated storage devices, creating discrete failure domains. This means that if one hard disk fails, only the corresponding core is affected, while other cores continue to operate normally with their own storage resources.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements local quality by giving each processor core its own dedicated I/O resources and memory space. This localized resource allocation ensures that storage operations on one core do not interfere with or propagate failures to other cores, thereby containing the failure domain to a single disk-core pair while maintaining overall system storage capacity.

Inventive Principle:
Principle #3Local quality

3Quantity of substance

If kernel images are shared across processing cores, then memory usage is reduced, but failure isolation is compromised

Engineering Contradiction:
Improvememory usageVSAvoidfailure isolation
Core Design Contradiction:
Quantity of substanceVSReliability

Solution Approach 1:

The patent allocates separate memory spaces to each processor core, with each core having its own dedicated kernel image. This segmentation prevents failure propagation between cores, as each core operates in an isolated memory environment. Although this increases total memory usage compared to sharing, it ensures that a failure in one core's kernel does not affect other cores, thereby achieving reliable failure isolation.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP2851807B1Method and system for supporting resource isolation under multi-core architecture
Publication Date: 2017.09.20 HUAWEI TECH CO LTD
  • EP2851807B1 patent drawingFigure 1a
  • EP2851807B1 patent drawingFigure 1b~2
  • EP2851807B1 patent drawingFigure 3

AI summary

Embodiments of the present invention provide a method and a system for supporting resource isolation in a multi-core architecture. In the method and system for supporting resource isolation in a multi-core architecture provided by the embodiments of the present invention, manners of inter-core operating system isolation, memory segment isolation, and I/O resource isolation are adopted, so that operating systems that run on different processing cores of the multi-core processor can run independently without affecting each other. Therefore, the present invention fully uses the advantages high integration level and low comprehensive costs of the multi-core processor, it is achieved that a failure domain of the multi-core processor remains in a single hard disk, and the multi-core processor has high reliability.