Fault Isolation via Ordered Fault Tables and Priority Logic
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fault detection and isolation methods in data processing systems face challenges in accurately identifying the source of failures due to error propagation, which can result in multiple fault reports with similar timestamps, and issues with timestamp accuracy for faults like clock and power failures, leading to inefficient analysis and potential human intervention.
Innovation Solution
An automated multi-step mechanism is implemented, using fault tables to order faults by processing flow, with a maintenance processor analyzing reports to identify the first-occurring fault within each unit, and employing priority values and timestamps to determine the primary source of failure, enabling automated action initiation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If timestamp method is used to isolate faults, then fault source identification is improved, but measurement precision deteriorates due to inaccurate timestamps for clock and power faults
Solution Approach 1:
The patent introduces a fault table as an intermediary data structure that maps fault types to their expected occurrence order in the processing flow. This mediator allows the system to determine fault causality relationships without relying solely on timestamps, thereby compensating for timestamp inaccuracy in clock and power faults while maintaining reliable fault isolation.
Solution Approach 2:
The patent segments the fault analysis process into multiple independent steps: collecting fault reports, searching fault tables for causal relationships, determining primary faults based on processing flow order, and isolating the fault source. This segmentation allows the system to handle different fault types appropriately, using timestamps when available but falling back to processing flow analysis when timestamps are inaccurate or unavailable.
2Reliability
If multiple fault indicators are used to detect faults, then fault detection coverage is improved, but device complexity increases due to ripple effect propagation
Solution Approach 1:
The patent divides the system into discrete units, each with its own fault table containing expected fault occurrence sequences. This segmentation allows the complex fault analysis problem to be broken down into manageable unit-level analyses, where the maintenance processor can systematically determine causal relationships by comparing fault reports against the segmented fault tables of individual units.
Solution Approach 2:
The patent implements a feedback mechanism where the maintenance processor collects fault reports from multiple indicators, analyzes them against the fault tables to determine causal relationships, and uses this analysis to identify the primary fault source. This feedback loop transforms the complex ripple effect of multiple activated indicators into a structured analytical process that systematically resolves the complexity.
3Productivity
If automated fault analysis is implemented, then productivity is improved, but device complexity increases due to maintenance processor requirements
Solution Approach 1:
The patent implements self-service by having the maintenance processor automatically perform fault analysis using pre-defined fault tables and processing flow sequences. The system autonomously collects fault reports, searches the fault tables for causal relationships, determines primary faults based on processing flow order, and identifies the fault source without requiring external human intervention or complex real-time analysis algorithms.
Solution Approach 2:
The patent applies preliminary action by pre-defining the fault tables and processing flow sequences before faults occur. This preparation allows the maintenance processor to perform rapid fault analysis when faults happen, simply by searching the pre-organized fault tables rather than performing complex real-time analysis, thereby improving productivity while keeping the added complexity manageable through static pre-computed data structures.
Data Source
AI summary
A mechanism for isolating failures in a digital system is provided. In one embodiment, a fault table is defined for each unit in the system. Related faults are ordered within the table to reflect the time-order in which the faults would be activated during operation of the associated unit. When multiple related faults are reported for a given unit in the system, the fault that is first located when a linear search of the corresponding fault table is conducted is considered the source of the failure within the unit. If faults are reported for multiple units, the source of the failure for the system is identified using at least one of primary and second priority values assigned to the faults, timestamps obtained when the faults are reported, and an order in which the faults are logged.