Cloud Resource Allocation via Predicted Deployment Growth
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional cloud computing systems face challenges in efficiently managing resource allocation, leading to deployment failures due to conservative policies that result in inefficient utilization of server resources and high computing costs, as they struggle to accommodate growing demands and maintain sufficient capacity buffers across node clusters.
Innovation Solution
A tenant growth prediction system that determines deployment growth classifications based on cluster features and utilization data, allowing for selective allocation of resources to maintain a capacity buffer threshold, thereby permitting or denying deployment requests to optimize resource utilization and prevent failures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conservative policies are used to maintain capacity buffers, then deployment failures are prevented, but resource utilization efficiency deteriorates and computing costs increase
Solution Approach 1:
The patent applies dynamics by transitioning from static conservative capacity buffer policies to dynamic policies that adapt to predicted deployment growth. The system continuously monitors cluster utilization and predicts future growth trends, then adjusts capacity buffer requirements accordingly. This allows the system to maintain reliability when growth is predicted while reducing unnecessary capacity buffers when growth is limited, thereby improving resource utilization efficiency.
Solution Approach 2:
The patent implements preliminary action by predicting future deployment growth before making capacity allocation decisions. The system analyzes historical utilization data and identifies growth trends in advance, allowing the capacity allocation engine to proactively adjust capacity buffers before deployment failures occur or resource waste accumulates. This predictive approach enables the system to maintain sufficient capacity when needed while avoiding excessive capacity allocation.
2Adaptability or versatility
If capacity is increased by adding new server nodes, then deployment requests are accommodated, but resource utilization efficiency deteriorates and computing costs increase
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the capacity buffer parameter based on predicted deployment growth. Instead of using fixed conservative capacity buffers or simply adding nodes to meet immediate demand, the system changes the capacity parameter according to growth predictions. When growth is predicted, the capacity buffer is increased; when growth is limited, the capacity buffer is reduced. This flexible parameter adjustment allows the system to accommodate deployment requests while maintaining high resource utilization efficiency.
3Reliability
If capacity buffers are maintained to accommodate growth, then deployment failures are prevented, but computing costs for providers and customers increase
Solution Approach 1:
The patent implements feedback by continuously monitoring actual cluster utilization and comparing it against predicted growth trends. The monitoring engine collects utilization data, the machine learning engine analyzes growth patterns, and the capacity allocation engine adjusts capacity buffers based on this feedback loop. This feedback mechanism ensures that capacity buffers are maintained only when necessary to prevent deployment failures, while unnecessary capacity allocation is eliminated, thereby reducing computing costs for both providers and customers.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure relates to systems, methods, and computer readable media for predicting deployment growth on one or more node clusters and selectively permitting deployment requests on a per cluster basis. For example, systems disclosed herein may apply tenant growth prediction system trained to output a deployment growth classification indicative of a predicted growth of deployments on a node cluster. The system disclosed herein may further utilize the deployment growth classification to determine whether a deployment request may be permitted while maintaining a sufficiently sized capacity buffer to avoid deployment failures for existing deployments previously implemented on the node cluster. By selectively permitting or denying deployments based on a variety of factors, the systems described herein can more efficiently utilize cluster resources on a per-cluster basis without causing a significant increase in deployment failures for existing customers.