Server Event Record Generation for IPMI Log Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
System administrators in IPMI-compliant computer systems face challenges in efficiently responding to unfamiliar sensor data records in the system event log, as they lack knowledge of all possible event records that could be recorded.
Innovation Solution
A method is introduced to generate a comprehensive list of possible event records by duplicating and modifying sensor data records, noting trigger events, and populating event-specific information for each copy, allowing for thorough analysis and management of potential system events.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the system administrator monitors the system event log without generating a comprehensive list of possible event records, then the operation remains simple, but the administrator cannot efficiently respond to unfamiliar sensor data records
Solution Approach 1:
The system pre-generates a comprehensive list of all possible event records by duplicating sensor data records and populating event-specific information before actual events occur. This preliminary action enables administrators to recognize and respond to unfamiliar events efficiently when they happen, as the complete event taxonomy is already established and available for reference.
2Adaptability or versatility
If the system generates and stores all possible event records for analysis, then event management capability is enhanced, but the storage and processing requirements increase
Solution Approach 1:
The system creates copies of sensor data records to generate all possible event records. By duplicating the base sensor data record structure and systematically varying the event-specific information fields, the system can enumerate all possible events without requiring storage of every actual event that occurs in practice. This copying approach provides complete event taxonomy coverage with manageable data volume.
3Reliability
If the system administrator lacks knowledge of all possible event records, then the system operation remains straightforward, but the administrator cannot properly respond to unfamiliar events
Solution Approach 1:
The system provides feedback to administrators by generating and presenting the complete list of possible event records along with their descriptions and characteristics. This feedback mechanism enables administrators to build knowledge of all possible events and improves their ability to recognize, understand, and respond accurately to unfamiliar events that occur in the system.
Data Source
AI summary
A system and method is disclosed in which the buses and physical characteristics of a server computer are monitored through server management software. A data structure for a monitored characteristic or group of characteristics is created and stored in a repository of data structures for other monitored devices and attributes within the server computer. As events, such as failure and pre-failure events, occur on one or more of the monitored attributes, the event is recorded in an event log. Using the server management software, monitoring commands can be issued by the baseboard management controller to each monitored bus to check the status of the attribute or device.