Cognitive Workload Classification in Multi-Tenant Cloud
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In cloud computing environments, accurately classifying variations in workload resource consumption across multiple tenants is challenging due to the dynamic nature of resource usage and the need to differentiate between normal fluctuations and anomalies, such as cyber attacks or new application deployments, which requires efficient data analysis and classification methods.
Innovation Solution
A cognitive classification system that analyzes both structured and unstructured data for tenants to determine classifications of workload resource consumption variations, using machine learning models to detect anomalies and predict potential causes, enabling cloud providers to take appropriate actions and optimize resource allocation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning models are used to classify workload variations, then classification accuracy is improved, but system complexity increases
Solution Approach 1:
The system segments the classification task by separating structured data analysis (using traditional statistical methods) from unstructured data analysis (using machine learning models). This allows the complex ML component to focus only on unstructured data, reducing overall system complexity while maintaining high accuracy through the combined approach
Solution Approach 2:
The patent introduces an intermediary layer that processes and prepares both structured and unstructured data before feeding them to the classification models. This intermediary processing layer simplifies the interaction between data sources and models, reducing system complexity while improving classification accuracy
2Measurement precision
If both structured and unstructured data are analyzed, then classification accuracy is improved, but data processing time increases
Solution Approach 1:
The system segments data processing by handling structured and unstructured data through separate pipelines. Structured data is processed quickly using efficient statistical methods, while unstructured data undergoes more intensive ML analysis. This segmentation allows parallel processing and reduces total processing time while maintaining comprehensive accuracy
Solution Approach 2:
The system performs preliminary processing and filtering on unstructured data before feeding it to machine learning models. This preliminary action reduces the volume of data that needs intensive analysis, thereby reducing overall processing time while preserving classification accuracy
3Reliability
If anomaly detection is performed to identify cyber attacks, then security is improved, but computational resources consumed increase
Solution Approach 1:
The system applies different quality levels of analysis to different data types. Unstructured data (which often contains security-relevant information) receives intensive ML-based analysis, while structured data receives lighter statistical processing. This local quality differentiation improves security detection effectiveness while reducing overall computational resource consumption
Solution Approach 2:
The system extracts and focuses ML analysis specifically on unstructured data portions that are most indicative of security anomalies. By taking out and prioritizing only the critical unstructured data for deep analysis, the system improves security detection while minimizing unnecessary computational resource consumption on already-processed structured data
Data Source
AI summary
One embodiment provides a method comprising receiving data relating to a tenant utilizing a cloud computing environment, and determining one or more classifications for a variation in current workload resource consumption of the tenant based on the data. The current workload resource consumption is indicative of current usage of one or more computing resources of the cloud computing environment.