Dynamic Computing Resource Relabeling for Cloud Scaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The significant scaling time required to acquire and configure computing resources in cloud computing systems can be longer than the time to execute a workload, due to the need for acquiring and reconfiguring resources for varying demands, which hampers efficient resource utilization.
Innovation Solution
Implementing a system that reuses allocated computing resources between workloads by identifying common resources and relabeling them for subsequent workloads, thereby reducing the need for duplicate resource acquisition and configuration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If computing resources are acquired and configured for each workload from cloud providers, then resource allocation flexibility is improved, but scaling time increases significantly
Solution Approach 1:
The system pre-acquires computing resources from cloud providers and maintains them in a ready state before workloads arrive. By performing the resource acquisition and configuration actions in advance, the system eliminates the need to wait for resources during workload execution, thus reducing scaling time while maintaining allocation flexibility
Solution Approach 2:
The system dynamically allocates and reconfigures pre-acquired computing resources based on varying workload demands. Resources can be quickly reassigned between different workloads through dynamic label updates and configuration changes, maintaining adaptability while avoiding repeated acquisition cycles that would increase scaling time
2Adaptability or versatility
If computing resources are reconfigured for varying workload demands, then workload support capability is improved, but scaling time increases
Solution Approach 1:
The system configures computing resources with universal, workload-agnostic settings that can serve multiple different workloads. By creating a standardized baseline configuration that satisfies common requirements across various workloads, the system reduces the need for extensive reconfiguration while maintaining the ability to support diverse workload types
Solution Approach 2:
The system efficiently reconfigures resources by making targeted parameter changes rather than complete reconfiguration. By identifying and modifying only the specific parameters that differ between workloads (such as labels, resource limits, or configuration flags) while keeping core settings unchanged, the system maintains workload support capability while minimizing scaling time
3Reliability
If resources are acquired and configured for each workload, then resource specificity is improved, but resource utilization efficiency decreases
Solution Approach 1:
The system merges multiple workload requirements into a shared pool of pre-acquired computing resources. By consolidating resources that can serve multiple workloads and managing them through a unified allocation system, the system maintains resource specificity through targeted assignment while improving overall utilization efficiency through shared ownership and reduced idle time
4Loss of energy
If cloud resources are acquired for brief periods of high demand, then cost efficiency is improved, but resource availability reliability decreases
Solution Approach 1:
The system pre-acquires computing resources during periods of lower demand when costs are lower and availability is higher. By securing resources in advance, the system ensures reliable resource availability during high-demand periods while taking advantage of more favorable pricing conditions during off-peak times, thus balancing cost efficiency with reliability
Data Source
AI summary
Some embodiments of the present disclosure are directed to systems, computer-readable media, and computer-implemented methods for dynamic computing resource management. Some embodiments are directed to identifying a computing resource common between a first workload and a second workload, replacing a label associated with the first workload on the identified computing resource with a label associated with the second workload, executing the second workload using the identified computing resource. Other embodiments may be disclosed or claimed.