PCI Adapter Boot Isolation via EEH

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In symmetric multiprocessor systems, faulty PCI adapters during the boot process cause significant delays and disruptions as they require identification, removal, and replacement, leading to prolonged boot times and system downtime.

Innovation Solution

Implementing Enhanced Error Handling (EEH) methods to actively disable defective PCI adapters during the boot process, allowing the system to continue booting without interruption, enabling 'hot swapping' of faulty devices without altering the machine state.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If a faulty PCI adapter is detected during boot process, then the boot process must be halted for identification and replacement, but this causes significant boot time delays and system downtime

Engineering Contradiction:
Improvesystem boot reliabilityVSAvoidboot time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent extracts the faulty PCI adapter from the boot process by detecting its presence and isolating it as a separate error condition. The EEH mechanism identifies the faulty adapter through error logs and isolates the specific PCI slot, allowing the boot process to continue without being blocked by the faulty device. This separates the fault detection and isolation functions from the main boot sequence.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent performs preliminary error detection and isolation actions during the boot process itself, rather than waiting for the faulty adapter to cause system failure. The EEH mechanism proactively checks for PCI errors and isolates faulty adapters before they can prevent system startup, enabling the system to proceed with booting even when faults are present.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If extensive testing of PCI adapters is performed during boot process, then system reliability is improved, but boot time is significantly increased

Engineering Contradiction:
Improvesystem component reliabilityVSAvoidboot time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent implements a mechanism to skip over faulty PCI adapters during the boot process. When the EEH mechanism detects a faulty adapter through error log analysis, it isolates the problematic slot and allows the boot process to continue without attempting to initialize or test the faulty device further, effectively rushing through the problematic area.

Inventive Principle:
Principle #21Skipping (Rushing through)

Solution Approach 2:

The patent performs partial testing of PCI adapters during boot by checking for error conditions rather than performing complete functional tests on each adapter. The EEH mechanism monitors for errors and isolates only those adapters that exhibit fault conditions, performing minimal necessary testing to ensure system reliability without exhaustive testing of all devices.

Inventive Principle:
Principle #16Partial or excessive action

3Measurement precision

If the system halts boot process to identify faulty PCI adapters, then accurate fault identification is achieved, but system availability is reduced

Engineering Contradiction:
Improvefault identification accuracyVSAvoidsystem availability
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent implements feedback mechanisms through EEH that continuously monitor PCI bus operations and error logs during the boot process. When errors are detected, the system provides feedback about the faulty adapter's identity and location, enabling automatic isolation without requiring manual intervention. This feedback loop maintains accurate fault identification while keeping the system operational.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The system performs self-service fault identification and isolation through the EEH mechanism. The boot process automatically detects faulty PCI adapters, identifies them through error log analysis, and isolates them without requiring external intervention or system halts. The system serves itself by handling fault identification and isolation internally, maintaining both accuracy and availability.

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS7650530B2Initializing a processing system to ensure fail-safe boot when faulty PCI adapters are present
Publication Date: 2010.01.19 X CORP
  • US7650530B2 patent drawing
  • US7650530B2 patent drawing
  • US7650530B2 patent drawing

AI summary

EEH methods are used during the boot process to actively disable a defective PCI adapter, thereby allowing the system boot to continue without disruption. This allows faulty adapters to be present in the machine without interrupting the boot process. The slots appear to be empty and the devices/adapters residing therein can be actively “hot swapped” out without altering the rest of the machine state.