Value Chain Workload Autoscaling for Cross-Enterprise Bottlenecks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cloud resource management systems fail to efficiently scale resources across enterprises in a value chain due to independent control of individual services, leading to bottlenecks and inefficiencies, particularly when one enterprise's workload fluctuates, affecting others without proactive resource allocation.
Innovation Solution
A cloud provider manages workload dependencies across enterprises by predicting and proactively allocating resources based on load-dependency relationships, using a prediction engine and load manager to scale resources automatically without sharing sensitive data, leveraging workload dependency data and SLAs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If cloud resources are independently controlled for each enterprise service, then each enterprise maintains data security and autonomy, but resource allocation efficiency deteriorates and bottlenecks occur when workload fluctuates
Solution Approach 1:
The cloud provider acts as an intermediary that receives workload information from one enterprise, predicts the impact on dependent services, and proactively allocates resources to affected services. This mediator approach enables coordinated resource management across enterprises without requiring them to share sensitive workload data directly, thus maintaining data security while improving resource allocation efficiency
Solution Approach 2:
The system performs preliminary resource allocation based on predicted workload fluctuations. When workload changes are anticipated for a service, resources are proactively allocated to dependent services before the actual workload impact occurs, preventing bottlenecks and improving overall resource allocation efficiency while maintaining enterprise autonomy
2Device complexity
If reactive resource scaling is used, then system complexity remains low, but service level agreement adherence deteriorates due to delays in responding to workload changes
Solution Approach 1:
The system implements preliminary resource scaling by predicting workload changes and allocating resources in advance before actual workload spikes occur. This proactive approach ensures service level agreement adherence by preventing resource bottlenecks before they impact service performance, while adding only moderate predictive complexity to the existing resource management system
3Productivity
If workload information is shared across enterprises, then resource allocation efficiency improves, but data security and privacy deteriorate
Solution Approach 1:
The cloud provider serves as a trusted intermediary that receives workload information from enterprises, processes it to predict impacts on dependent services, and triggers resource allocation without requiring enterprises to share sensitive workload data with each other. This maintains data security and privacy while enabling efficient cross-enterprise resource allocation through the intermediary's predictive analytics
Solution Approach 2:
The system creates and uses predictive models that replicate workload patterns and dependencies without transferring actual sensitive workload data between enterprises. By working with anonymized patterns and predictions rather than raw data copies, the system achieves resource allocation efficiency while preserving data security and enterprise confidentiality
Data Source
AI summary
The technology described herein is directed towards automatically scaling cloud provider resources allocated for one enterprise's service based on the resources being used or expected to be used by another enterprise's service, in which there is a value chain-based load-dependency relationship between the two enterprises' services. In one example implementation, a first prediction engine determines a predicted workload for a first enterprise service, and sends that information to a load manager that allocates resources for the first enterprise service based on the prediction. Based on the load-dependency relationship, the first prediction engine sends the predicted workload to a second prediction engine, which predicts a second predicted workload for a second enterprise's service, and sends the second prediction information to a load manager that allocates resources for the second enterprise service based on the second prediction. The automatic scaling is done without sharing any enterprise-sensitive data among the enterprises.