Data Center Fault Diagnosis Using Dependency-Aware Log Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fault diagnosis methods in data centers are inefficient and time-consuming due to administrators' lack of awareness of deployment and dependency relationships, especially in distributed centers with geographically dispersed machines, leading to inaccurate or lengthy diagnosis processes.
Innovation Solution
A method that monitors fault incidents, determines the business logic executed at the affected node, selects relevant log files based on a management model reflecting deployment and application dependencies, and diagnoses the fault using a processor to identify error log entries and determine the root cause.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If administrators manually check each application and service using test check methodology, then they can identify fault causes, but the diagnosis process becomes time-consuming and inefficient
Solution Approach 1:
The system performs preliminary actions by automatically collecting and organizing deployment information, dependency relationships, and log data before faults occur. When a fault happens, the diagnosis system can immediately query pre-prepared information about application dependencies and deployment configurations, eliminating the need for administrators to manually gather this information during fault diagnosis.
Solution Approach 2:
The system introduces an intermediary fault diagnosis system that acts as a mediator between administrators and complex deployment information. This intermediary automatically queries deployment databases, analyzes log files, and presents simplified fault information to administrators, reducing the time and effort required for manual investigation.
2Measurement precision
If administrators manually investigate fault incidents, then they can determine fault causes, but the process becomes complex and error-prone due to lack of awareness of deployment and dependency relationships
Solution Approach 1:
The system introduces an intermediary fault diagnosis system that acts as a mediator between administrators and complex deployment information. This intermediary automatically queries deployment databases, analyzes log files, and presents simplified fault information to administrators, reducing the time and effort required for manual investigation.
Solution Approach 2:
The system implements feedback mechanisms where the fault diagnosis system continuously monitors system state, compares actual behavior against expected behavior based on dependency relationships, and provides feedback about anomalies. This automated feedback loop reduces the complexity of manual investigation by highlighting only the relevant dependencies and configurations that may be causing faults.
3Adaptability or versatility
If the data center topology changes with new virtual machines, then the system becomes more flexible and adaptable, but manual diagnosis becomes more error-prone and less reliable
Solution Approach 1:
The system embraces dynamics by automatically adapting to topology changes. When new virtual machines or services are deployed, the system dynamically updates its internal models of deployment relationships and dependencies. This dynamic adaptation ensures that fault diagnosis remains reliable even as the data center evolves, eliminating the need for manual updates to diagnostic knowledge bases.
4Quantity of substance
If physical machines are distributed at different geographical locations, then the data center capacity increases, but the time required for fault diagnosis becomes lengthy
Solution Approach 1:
The system introduces an intermediary fault diagnosis system that acts as a mediator between administrators and complex deployment information. This intermediary automatically queries deployment databases, analyzes log files, and presents simplified fault information to administrators, reducing the time and effort required for manual investigation.
Solution Approach 2:
The system performs preliminary actions by automatically collecting and organizing deployment information, dependency relationships, and log data before faults occur. When a fault happens, the diagnosis system can immediately query pre-prepared information about application dependencies and deployment configurations, eliminating the need for administrators to manually gather this information during fault diagnosis.
Data Source
AI summary
A method, computer program product and apparatus for diagnosing a fault incident in a data center. A determination is made as to whether a fault incident happened in the data center. If a fault incident happened, then the business logic being executed in a node where the fault incident happened is determined. The log files associated with the execution process of the business logic is selected based on a management model corresponding to the data center, where the management model reflects the deployment and application dependency of the business logics. The fault incident is diagnosed based on the log files. The management model is utilized to select the log files associated with the execution process of the business logic, and to write information of the business logic and dependency relationships between the applications into the log file, so as to automatically, rapidly and accurately diagnose the fault incident.