Error Isolation Unit for PCIe Link Error Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Vertically integrated computer systems face challenges in distinguishing between power supply errors and other hardware errors, particularly when the power supply units for the server and storage apparatuses are separate, leading to incorrect error detection and complex manual procedures for error handling.
Innovation Solution
A computer system with an error isolation unit that periodically checks error status registers and link status registers to differentiate between power supply errors and other errors, masking interruptions for power supply errors and detecting them through polling, thereby preventing incorrect error reporting and allowing for automatic reactivation of devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of energy
If the power supply unit for the device is stopped, then the device can be powered down to save energy or for maintenance, but the server apparatus or storage apparatus detects it as a hardware error of the device
Solution Approach 1:
The patent introduces an intermediary mechanism (error isolation unit/software module) that mediates between the power supply unit and the error detection system. This intermediary checks the power supply state before triggering error detection, preventing false positives when power is intentionally stopped. The intermediary layer filters out spurious error signals caused by power shutdown from actual hardware failures.
Solution Approach 2:
The patent changes the parameter used for error detection by introducing a power supply state check as a preliminary condition. Instead of directly detecting hardware errors without context, the system now evaluates multiple parameters including power supply status, link status, and error type before concluding a hardware failure. This multi-parameter approach prevents misinterpretation of power shutdown as hardware error.
2Ease of operation
If hot-remove operations are performed before stopping the power unit, then the device can be properly detached from the system, but complex manual procedures are required and users may make false operations
Solution Approach 1:
The patent implements self-service functionality where the system automatically handles the error isolation process without requiring manual user intervention. When power shutdown is detected, the error isolation unit automatically checks conditions, determines whether isolation is needed, and executes the appropriate error handling procedures autonomously. This eliminates the need for users to perform complex hot-remove operations while preventing false operations.
Solution Approach 2:
The patent performs preliminary actions by checking the power supply state and link status before finalizing error detection. The system proactively verifies whether the device should be isolated based on predetermined conditions (power supply status, link status, error type) before triggering full error handling procedures. This preliminary check prevents unnecessary complex operations and guides the system through appropriate handling steps automatically.
3Stability of the object's composition
If the PCIe link is disconnected when power supply unit stops, then the device is properly isolated from the system, but the storage apparatus wrongly detects it as an error of the NTB device
Solution Approach 1:
The patent implements feedback mechanisms where the error isolation unit continuously monitors the power supply state and link status, using this feedback information to determine whether a detected error is actually caused by power shutdown. The system compares current error conditions with the power supply state feedback to distinguish between intentional power shutdown (stable configuration) and actual hardware errors (precise detection). This feedback loop prevents wrong detection while maintaining system stability.
Solution Approach 2:
The patent segments the error detection process into distinct phases: power supply state check, link status verification, error type identification, and isolation determination. By dividing the error detection function into separate checkable components, the system can precisely evaluate each condition independently. This segmentation allows the system to identify whether the PCIe link disconnection is due to power shutdown or actual device error, improving measurement precision without compromising system stability.
Data Source
AI summary
A computer system comprises a computer being coupled to a device via a device interface. The device interface has an error status register and a link status register. An error isolation unit periodically obtains values of the error status register and the link status register. The error isolation unit determine whether an error occurs in the device; determine whether the error is an error to be isolated; determine whether the error is an error to be isolated based on values of the error status register and the link status register re-obtained after elapse of a predetermined time in a case of determining the error is not to be isolated. The error isolation unit detects the error as an error of a protocol caused by stop of a power supply in a case of determining the error is to be isolated.