Cloud Capacity Resolver for Pruning Redundant VM Migration Actions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data centers face challenges with over-provisioning of virtual machines (VMs), leading to inefficient resource utilization and burdensome request handling, which can result in error-prone and non-deterministic operations.
Innovation Solution
A capacity resolver system that includes a cloud orchestration server and hypervisors, which determines if parameters exceed thresholds, prioritizing downsizing underutilized resources over migration to optimize VM allocation, thereby improving resource utilization and reducing over-provisioning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If manual capacity management processes are used, then flexibility in handling complex boundary conditions is maintained, but processing time and operational complexity increase significantly
Solution Approach 1:
The patent replaces manual mechanical capacity management processes with an automated machine learning-based system. The ML model automatically analyzes capacity metrics, predicts future capacity needs, and makes provisioning decisions without human intervention, thereby reducing processing time while maintaining optimal capacity management through data-driven insights
2Measurement precision
If more comprehensive capacity metrics are collected, then provisioning accuracy improves, but system complexity and data processing requirements increase
Solution Approach 1:
The system automatically collects, processes, and analyzes capacity metrics from multiple sources without requiring manual configuration or complex data processing infrastructure. The ML model self-tunes to identify relevant metrics and patterns, reducing system complexity while maintaining high provisioning accuracy through automated adaptive learning
3Loss of energy
If VM allocation is optimized to increase controller and memory utilization, then cost efficiency improves, but allocation complexity and risk of over-provisioning increase
Solution Approach 1:
The patent implements continuous feedback loops where the ML model monitors actual VM performance and capacity utilization in real-time. This feedback enables dynamic adjustment of allocations to optimize resource usage while preventing over-provisioning through data-driven decisions based on actual workload patterns and capacity trends
Data Source
AI summary
A capacity resolver system in a cloud-based multi-tenant system includes point of presence (POP) systems and a cloud orchestration server. The POPs include hypervisors and the hypervisors includes nodes. A request for provisioning a node in a POP is received. Parameters are received from the hypervisors of the POP. Triggering of one or more parameters above respective threshold values is determined. Downsizing or migration of one or more nodes of the plurality of nodes is selected based on the triggering of the one or more parameters is selected. Based on the selection of downsizing or migration of the one or more nodes of the plurality of nodes, the plurality of actions are evaluated iteratively such that one or more actions are excludable or replaceable with an action. Based on this evaluation, the one or more actions are pruned without over-provisioning the plurality of nodes.