Root Cause Identification for Virtualized Application Underachievement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In a virtually provisioned environment, identifying the root cause of application underachievement in meeting service level agreements is challenging due to the complexity of processing operations across thousands of servers, making it difficult to prevent or quickly resolve violations.
Innovation Solution
An analysis application records utilization measurements for CPU, memory, network, and I/O resources for each server when an operation exceeds a time constraint, converts these measurements into utilization ranges, and uses an apriori algorithm to identify specific resources and servers that consistently contribute to underperformance, thereby pinpointing the root cause.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If resource utilization measurements are collected for all servers when time constraints are exceeded, then root cause identification accuracy is improved, but system complexity and data processing load increase
Solution Approach 1:
The patent segments the vast server environment into manageable groups by organizing servers according to their roles and relationships in processing application operations. This segmentation allows the system to focus measurements on relevant server groups rather than all servers uniformly, reducing data processing complexity while maintaining root cause identification accuracy.
Solution Approach 2:
The patent transforms raw resource utilization measurements into standardized utilization ranges, converting continuous parameter values into discrete categories. This parameter transformation simplifies the analysis of resource utilization patterns across thousands of servers while preserving the ability to identify root causes of service level agreement violations.
2Difficulty of detecting and measuring
If comprehensive resource monitoring is implemented across thousands of servers, then detection capability is improved, but processing time and computational resources increase
Solution Approach 1:
The patent performs preliminary organization of servers into structured groups based on their roles and relationships before analyzing resource utilization data. This preliminary structuring enables faster querying and analysis when time constraints are exceeded, reducing the processing time required to identify root causes while maintaining comprehensive detection capability across the server environment.
3Manufacturing precision
If detailed utilization measurements are recorded for each server, then analysis precision is improved, but data storage requirements and processing overhead increase
Solution Approach 1:
The patent converts detailed continuous utilization measurements into discrete utilization ranges for each resource type and server. This parameter discretization reduces the volume of data that must be stored and processed while preserving sufficient precision to identify root causes of application underachievement through pattern recognition in the binned data.
Data Source
AI summary
Identifying root causes of application underachievement in a virtually provisioned environment is described. A utilization measurement is recorded for each resource for each server in a virtually provisioned environment associated with an application operation if an application operation time exceeds a time constraint. The resources include a central processing unit resource, a memory resource, a network resource, and/or an input-output resource. Each utilization measurement is converted to a corresponding utilization range of multiple utilization ranges. After the application operation time exceeds the time constraint on a specified number of occasions, an identification of a specific resource for a specific server as a root cause of the application operation time exceeding the time constraint is output if the utilization record includes a specific range for the specific resource for the specific server more than a specified frequency.