Application Command Adjustment Using Host-Specific Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern computing systems face challenges in executing application commands across different hosts due to varying formats and syntaxes, making it difficult to determine how to execute commands effectively without manual parsing of logs and potential human intervention.
Innovation Solution
Utilizing machine learning models to identify and resolve values for dynamic variables in application commands, allowing commands to be executed correctly on hosts without prior discovery, by generating a list of values to replace temporary variables.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual parsing of logs is used to determine command execution values, then accuracy of variable identification is improved, but labor intensity and time consumption increase
Solution Approach 1:
The patent replaces manual mechanical parsing of logs with an automated machine learning system. The ML model automatically extracts and identifies variable values from application commands and host information, eliminating the need for manual log analysis while maintaining high accuracy in variable identification.
Solution Approach 2:
The system enables self-service automation where the machine learning model autonomously performs variable identification without human intervention. The model takes application commands and host details as input and automatically generates the necessary variable values for command execution, making the system self-sufficient.
2Reliability
If discovery process is run on each host to identify variable values, then reliability of command execution is improved, but productivity decreases
Solution Approach 1:
The patent implements preliminary action by pre-training a machine learning model on discovery data from multiple hosts. Instead of running discovery on each individual host, the system performs discovery once during model training, then uses the trained model to rapidly identify variable values for any host, maintaining reliability while dramatically improving productivity.
Solution Approach 2:
The machine learning model serves as a universal solution that can identify variable values across different hosts and applications. The single trained model handles multiple hosts without requiring separate discovery processes for each, making the system universally applicable and highly productive.
3Ease of operation
If temporary values are used for dynamic variables in commands, then ease of command generation is improved, but execution accuracy deteriorates
Solution Approach 1:
The patent changes the parameter state of variable values by using the machine learning model to transform temporary placeholder values into accurate host-specific values. The model takes temporary values as input and outputs corrected, execution-ready values that maintain both ease of command generation and high execution accuracy.
Data Source
AI summary
Techniques described herein relate to using one or more machine learning models to identify values of variables to be used in application commands for execution. For example, the one or more machine learning models may identify one or more values of variables that are used to replace a different set of one or more values of variables in an application command before executing the command on a host.