Log File Collection System for Failure Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In large systems, it is difficult for system operators to grasp all log files, especially those with low reference frequency, making it challenging to determine appropriate log files for failure analysis when an alert occurs, as existing systems require predefined storage destinations which are hard to define due to the complexity and size of the systems.
Innovation Solution
A file information collection system that acquires process IDs and file attributes from servers, generates a list of relevant log files based on predetermined conditions, and prioritizes files by update time to identify appropriate log files for analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If a system operator tries to grasp every log file in a large system, then complete information is obtained, but the time and effort required becomes unmanageable
Solution Approach 1:
The system extracts only the necessary log files related to the alert from the entire system. When an alert occurs, the notification device identifies and extracts log files from specific processes (caller and called processes) rather than requiring operators to examine all log files in the system, thus obtaining complete relevant information without the time cost of reviewing everything.
Solution Approach 2:
The system segments the large system into individual processes and identifies which specific processes are related to the alert. By dividing the system into manageable process units and selectively examining only those involved in the alert (through parent-child process relationships), the operator can grasp necessary log file information without being overwhelmed by the entire system's log files.
2Ease of manufacture
If the system requires predefined storage destinations for log files, then log collection is simplified, but the system complexity increases due to the need to define storage destinations for each log file
Solution Approach 1:
The system performs self-service by automatically identifying log file storage destinations through process information rather than requiring predefined configuration. The notification device acquires process information from the alert notification and automatically determines which log files are relevant, eliminating the need for operators to manually define storage destinations for each log file while maintaining simplified log collection.
3Loss of information
If all log files are collected for failure analysis, then complete diagnostic information is available, but the difficulty of identifying relevant files increases
Solution Approach 1:
The system uses feedback from the alert notification to identify relevant log files. The notification device receives alert information, uses this feedback to determine which processes are involved in the failure, and then identifies log files specifically from those processes. This feedback mechanism ensures complete diagnostic information is available while making identification straightforward, as the alert itself provides the criteria for selection.
Solution Approach 2:
The system introduces process information as an intermediary between the alert and the log files. Rather than directly linking alerts to log files or requiring operators to search through all log files, the process information serves as a mediator that connects the alert notification to the specific log files that contain relevant diagnostic information, simplifying the identification process while ensuring completeness.
Data Source
AI summary
Provided is a file information collection system etc., to determine an appropriate log file for failure analysis in a case that an alert occurs in any server belonging to a system in operation and to generate a list of a file name of the file. In the system includes; information acquisition means that acquires a process ID of each process etc., used by the each process; list generation means that selects a file whose name meets a predetermined condition, generates a list of a record including identification information of a server, and, adds the list at a bottom of an already-generated list; and sequence control means that selects the record from a top of the list, acquires information about an update time of a file, and, changes a server attribute in the selected record and moves the record to a bottom of the list.