Load Characteristic Estimation via Program Feature Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Programmers face challenges in predicting the computer resources required by their programs, especially in shared large-scale distributed systems, as existing methods require actual program execution or are limited by language-specific analysis, making it difficult to estimate load characteristics accurately without running the program.
Innovation Solution
A load characteristic estimation system that inputs a program, extracts features, and estimates load characteristics by reading similar program data from a recording unit, allowing for prediction without executing the program, using a similarity search and load characteristic computation unit.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If static analysis is used to estimate execution time, then analysis can be performed without running the program, but the analysis is limited by language-specific constraints and cannot accurately predict load characteristics
Solution Approach 1:
The patent creates feature vectors that copy and represent the essential characteristics of programs in a language-agnostic manner. By extracting semantic features and transforming them into universal vector representations, the system can analyze programs from different programming languages using the same methodology, overcoming the language-specific limitations of traditional static analysis.
Solution Approach 2:
The patent transforms program characteristics from their original language-specific forms into standardized feature vectors with consistent parameters. This parameter transformation allows the system to maintain accuracy in execution time estimation while gaining versatility across different programming languages through uniform feature representation.
2Measurement precision
If actual program execution is performed to measure load characteristics, then accurate resource requirements are obtained, but programmers cannot predict resource needs before deployment in shared systems
Solution Approach 1:
The patent performs preliminary analysis by extracting feature vectors from program source code before actual execution. This preliminary action captures the essential characteristics of the program, enabling the system to predict load characteristics and resource requirements in advance, allowing programmers to plan resource allocation before deploying to shared computing systems.
Solution Approach 2:
The patent introduces feature vectors as an intermediary between program source code and load characteristic measurement. These vectors serve as a bridge that captures program characteristics without requiring actual execution, enabling accurate prediction of resource needs while maintaining the ability to validate predictions through subsequent actual measurements when performed.
3Ease of operation
If similar programs are used to estimate load characteristics, then resource prediction is enabled without execution, but the system requires a database of known program characteristics
Solution Approach 1:
The patent creates compressed feature vector representations of programs that capture essential characteristics in a standardized format. By copying and storing these compact feature vectors rather than complete program databases, the system enables efficient similarity comparison and load characteristic estimation while reducing the complexity and storage requirements of the reference database.
Data Source
AI summary
A load characteristic estimation system includes: a program input unit into which a program that runs on a prescribed computer system is inputted; a program feature information acquisition unit that obtains features from the inputted program; and a load characteristic estimation unit that estimates a load characteristic of the inputted program by reading load characteristic information of another program having a feature similar to the inputted program, from a load characteristic recording unit that records a load characteristic produced in the prescribed computer system when the other program was run on the computer system in the past.