Resource Management Platform for Cluster Provisioning Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cluster-based services incur significant costs due to hourly billing, even for short usage periods, and lack efficient means for shared resource provisioning among users or groups, leading to inefficient resource utilization and increased costs.
Innovation Solution
A management platform that allows users or groups to define execution constraints, dynamically provisioning and terminating clusters or instances based on cost, time, performance, and size limits, optimizing resource allocation and usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If hourly billing is used for cluster-based services, then service providers can simplify billing calculations, but users incur significant costs even for short usage periods
Solution Approach 1:
The patent changes the billing parameter from fixed hourly rates to usage-based metrics (e.g., per GB processed, per task completed). This allows billing to reflect actual resource consumption rather than time-based allocations, reducing user costs for short tasks while maintaining simple calculation methods based on measurable output parameters.
Solution Approach 2:
The system automatically monitors and tracks actual resource usage metrics (data processed, tasks completed) and generates billing based on these self-reported measurements. This eliminates the need for manual time-tracking while ensuring users only pay for actual usage, resolving the contradiction between billing simplicity and cost accuracy.
2Productivity
If clusters are provisioned for group sharing, then resource utilization efficiency increases, but provisioning management complexity increases
Solution Approach 1:
The patent introduces a resource manager intermediary that handles all provisioning decisions between users and physical resources. This manager automatically allocates clusters to group members based on predefined policies and current utilization, eliminating the need for complex manual coordination while maximizing resource sharing efficiency.
Solution Approach 2:
The system dynamically adjusts cluster provisioning based on real-time group usage patterns and resource availability. Rather than static allocations, the resource manager continuously optimizes cluster distribution to group members, adapting to changing conditions without requiring complex manual reconfiguration.
3Loss of energy
If execution constraints are processed to determine maximum clusters, then cost control improves, but processing complexity increases
Solution Approach 1:
The patent processes execution constraints and determines maximum cluster allocations in advance, before actual resource provisioning occurs. By pre-calculating cost-based limits and availability constraints, the system simplifies real-time provisioning decisions while maintaining strict cost control, avoiding complex calculations during active resource usage.
Data Source
AI summary
An approach is provided for managing the provisioning and utilization of resources. A management platform determines a request from a user for execution of one or more data processing tasks by a remote computing service. The management platform also processes and/or facilitates a processing of at least one execution constraint associated with the user, a group associated with the user, or a combination thereof to determine a maximum number of clusters, cluster instances, or a combination thereof of the remote computing service to be provisioned for fulfilling the request. The management platform further causes, at least in part, a provisioning of one or more clusters, one or more cluster instances, or a combination thereof to the user, the group, or a combination thereof to within the maximum number of clusters, cluster instances, or a combination thereof based on the at least one execution constraint.