Unified BIOS BMC Error Log Collection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for collecting error logs in servers are inadequate, as they lack detailed error detection capabilities, particularly for hardware errors, and require on-site administration for error analysis, making it difficult to diagnose and repair issues efficiently.
Innovation Solution
A method that involves the CPU generating structured BIOS error logs during the booting process, which are transmitted to the BMC for packaging with BMC sensor logs to create a comprehensive error log file, allowing for remote analysis and storage, facilitating easier error detection and repair.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If BMC or BIOS generates SEL for error detection, then error detection capability is provided, but the error logs do not contain sufficient details to find the cause of hardware errors
Solution Approach 1:
The patent combines BIOS error logs (containing hardware error information from POST) and BMC system event logs into a single unified error log file. This merging ensures that both the error detection capability of BMC/BIOS and the detailed hardware error information from BIOS are preserved together, resolving the contradiction between providing error detection and maintaining error detail information.
2Measurement precision
If specific error detection firmware is used to generate error detection logs, then error logs can be generated, but an administrator has to be on site to connect the server to external hardware device, making the process time consuming and inconvenient
Solution Approach 1:
The patent enables the server to automatically generate and store error log files by integrating BIOS and BMC error logging capabilities. The server performs error detection and log generation autonomously without requiring external hardware devices or on-site administrator intervention, thus resolving the contradiction between maintaining error detection capability and reducing time loss.
3Measurement precision
If specific error detection firmware is used to generate error detection logs, then error logs can be generated, but the logs may not completely reflect the condition under which the error occurred, making subsequent analysis and repair difficult
Solution Approach 1:
The patent merges BIOS error logs (which capture hardware error conditions during POST) with BMC system event logs into a comprehensive error log file. This combination ensures that the error logs completely and accurately reflect the error conditions, including both the error detection information from BMC and the detailed hardware error context from BIOS, thereby resolving the contradiction between error detection capability and accuracy of error condition reflection.
4Adaptability or versatility
If BMC is used for managing operation and monitoring health of the server, then system management is provided, but the types and detail levels of errors detectable by the BMC are limited
Solution Approach 1:
The patent combines the system management capability of BMC with the detailed hardware error detection capability of BIOS. By merging the error logs from both sources, the system maintains BMC's versatile management function while supplementing it with BIOS's detailed hardware error detection, thereby resolving the contradiction between system management adaptability and error detection detail level.
Data Source
AI summary
A method of collecting error logs according to the disclosure includes generating, during procedure of BIOS of a server, at least one BIOS error log based on detection of an error condition of one or more of hardware devices and a CPU, transmitting the at least one BIOS error log to a BMC, storing the at least one BIOS error log received from the CPU, packaging the at least one BIOS error log and at least one log that is generated by the BMC and that is related to BMC sensors to generate an error log file, and storing the error log file.