Multi-Workload Identification Using SOCC in Heterogeneous Data Centers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods are inadequate for identifying multiple workloads simultaneously in a heterogeneous environment, failing to efficiently allocate resources and optimize performance due to their focus on single-workload identification.
Innovation Solution
A method and system utilizing a Successively Ordered Classifier Chain (SOCC) module to identify multiple workloads by collecting and correlating parameters from various system stack layers, creating a feature set, and processing it to enhance workload recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing single-workload identification methods are used, then the system can identify one workload at a time, but it fails to identify multiple workloads simultaneously leading to inefficient resource allocation
Solution Approach 1:
The system segments the workload identification process into distinct classification stages using a Successively Ordered Classifier Chain (SOCC) module. Each classifier in the chain handles specific workload types, enabling simultaneous identification of multiple workloads by dividing the complex identification task into manageable segments that process different workload characteristics independently and sequentially.
Solution Approach 2:
The patent introduces a new dimension to workload identification by implementing a multi-layer classification architecture. The SOCC module adds a temporal dimension through successive classification stages, where each stage processes workload parameters in a specific sequence, enabling the system to distinguish multiple workloads simultaneously by analyzing them across different classification dimensions rather than treating them as a single homogeneous group.
2Adaptability or versatility
If the system attempts to identify multiple workloads simultaneously using complex methods, then multi-workload recognition capability improves, but system complexity and computational overhead increase
Solution Approach 1:
The complex task of multi-workload identification is segmented into a chain of simpler binary classifiers. Each classifier in the SOCC module handles a specific decision point, breaking down the complex multi-class classification problem into a sequence of manageable binary decisions. This segmentation reduces the computational complexity of each individual classifier while maintaining the overall capability to identify multiple workloads simultaneously.
Solution Approach 2:
The system employs partial classification action by implementing a successively ordered classifier chain where not all classifiers need to be executed for every workload identification task. The chain allows early termination when sufficient classification accuracy is achieved, performing only the necessary portion of the classification sequence required to distinguish the workloads present, thereby reducing overall computational overhead while maintaining multi-workload identification capability.
3Measurement precision
If comprehensive parameters are collected from multiple system stack layers, then workload identification accuracy improves, but data processing time and computational resources increase
Solution Approach 1:
The system performs preliminary action by pre-processing and organizing parameters from multiple system stack layers before they reach the SOCC module. Parameters are collected, correlated, and structured in advance during normal system operation, allowing the classification process to work with pre-organized data rather than raw parameters. This preliminary organization of data from different system layers reduces the processing time required during actual workload identification while maintaining comprehensive parameter analysis for accurate identification.
Solution Approach 2:
The parameter collection and processing workflow is segmented into distinct stages: parameter collection from multiple system stack layers, parameter correlation and feature extraction, and final classification by the SOCC module. This segmentation allows each stage to be optimized independently, with parameter collection occurring asynchronously during system operation and classification performed only when sufficient data is available, reducing overall processing time while maintaining comprehensive analysis for accurate multi-workload identification.
Data Source
AI summary
A method for identifying a plurality of workloads in a heterogenous environment includes collecting a plurality of parameters from at least one layer of a system stack associated with the plurality of workloads, correlating the collected plurality of parameters from different layers of software stack, and creating a feature set based on the correlated plurality of parameters. The method further includes processing the feature set using a successively ordered classifier chain (SOCC) module to identify the presence of the plurality of workloads in the heterogenous environment in a data center.