Selective Error Recovery Using Privilege-Level Discrimination

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In complex multi-processor systems, error recovery is cumbersome due to the lack of a central protection mechanism, leading to prolonged downtime and network disruptions when all execution threads and their configurations need to be reloaded following an error.

Innovation Solution

Implementing a privilege level system where processing units can determine the severity of errors and respond accordingly, allowing for selective reset of threads and resources based on privilege levels, reducing the need for full system reloads.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If all execution threads for all processors are reloaded together with their entire configuration following an error, then system reliability is improved through comprehensive error recovery, but device downtime increases and network performance deteriorates

Engineering Contradiction:
Improveerror recoveryVSAvoiddevice downtime
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent segments the error recovery process by introducing privilege levels that divide the system into hierarchical groups. When an error occurs, only threads at or below the error's privilege level are reloaded, rather than reloading all threads across all processors. This segmentation allows selective recovery of affected components while maintaining operational threads, thereby reducing downtime while ensuring reliable error handling for the impacted segments.

Inventive Principle:
Principle #1Segmentation

2Reliability

If all execution threads for all processors are reloaded together with their entire configuration, then complete system state restoration is achieved, but network availability decreases and other devices experience slowdowns

Engineering Contradiction:
Improvesystem state restorationVSAvoidnetwork availability
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent applies local quality by making error recovery scope dependent on the local privilege level of the error. Each error is assigned a privilege level that determines its local impact zone. Threads operating at higher privilege levels are preserved unless directly affected, while only threads at or below the error's privilege level are reloaded. This localized approach restores necessary system state for affected areas while maintaining network availability through continued operation of unaffected high-privilege threads.

Inventive Principle:
Principle #3Local quality

3Ease of operation

If a centralized error protection mechanism is used in single processor systems, then error control is simplified and more effective, but this approach cannot be directly applied to multiple processor systems without increasing complexity

Engineering Contradiction:
Improveerror controlVSAvoiderror recovery complexity
Core Design Contradiction:
Ease of operationVSDevice complexity

Solution Approach 1:

The patent creates a universal error handling framework that works across multiple processors by introducing privilege levels as a common organizing principle. The privilege level system serves multiple functions: it classifies errors, determines recovery scope, and identifies affected threads across different processors. This multi-functional approach provides centralized-like error control in distributed systems without adding proportional complexity, as the same privilege level mechanism applies uniformly across all processors regardless of system size.

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

Data Source

PatentUS7721151B2Selective error recovery of processing complex using privilege-level error discrimination
Publication Date: 2010.05.18 CISCO TECHNOLOGY INC
  • US7721151B2 patent drawing
  • US7721151B2 patent drawing
  • US7721151B2 patent drawing

AI summary

An apparatus has at least one processing unit to generate a request having a request privilege level. At least one resource exists in the apparatus to receive the request and determine if the request is allowable. The apparatus includes an error handler that determines the nature of an error and performs a reset based upon the privilege level of the request that cause the error.