Auto-scaling Host Machines for VDI Capacity Balancing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing VDI/DaaS systems face challenges in configuring auto-scaling policies to balance user experience and costs effectively, as existing schedule-based and load-based policies often result in suboptimal user experience or higher costs.
Innovation Solution
A new 'balanced auto-scale policy' that uses a single configuration value (P) to set the probability of available capacity for user sessions, generating statistical models from historical data to dynamically adjust host machine capacity based on user demand and power-on times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If excess capacity is maintained to ensure immediate user access, then user experience is improved, but costs increase
Solution Approach 1:
The system dynamically adjusts host machine capacity based on real-time and historical demand patterns. Instead of maintaining static excess capacity, the system continuously scales resources up or down to match actual user needs, ensuring reliable user experience while minimizing wasted capacity and associated costs.
Solution Approach 2:
The system uses historical data to predict future demand patterns and proactively provisions capacity in advance. By analyzing past usage trends, the system prepares appropriate capacity before peak demand occurs, ensuring user experience is maintained without requiring permanent excess capacity that would increase costs.
2Loss of energy
If capacity is reduced to minimize costs, then cost savings are achieved, but user experience deteriorates due to waiting time
Solution Approach 1:
The system continuously monitors actual user demand and power-on performance, using this feedback to refine its capacity management decisions. By tracking historical data about demand patterns and power-on times, the system learns from past performance and adjusts capacity allocation to prevent both waste and user waiting, optimizing the balance between cost and experience.
Solution Approach 2:
The system autonomously manages capacity allocation without requiring manual intervention. It automatically scales resources based on predicted and actual demand, making real-time decisions about when to provision or deprovision capacity, thereby eliminating the need for administrators to manually balance cost versus user experience.
3Adaptability or versatility
If multiple configuration parameters are exposed for auto-scaling, then fine-grained control is achieved, but system complexity increases
Solution Approach 1:
The system extracts and automates the complex multi-parameter configuration process. Instead of exposing numerous individual settings for administrators to tune, the system internally manages multiple parameters (demand patterns, power-on times, capacity thresholds) through automated algorithms, presenting a simplified interface while maintaining fine-grained control capabilities.
Solution Approach 2:
The system performs self-configuration by automatically determining optimal capacity settings based on historical data and demand patterns. The automated system replaces manual administrator configuration, eliminating the complexity of multiple settings while maintaining adaptability through data-driven decision-making.
Data Source
AI summary
According to one aspect, a method can include: receiving, by a computing device, historical data for an organization having a plurality of host machines that can be selectively powered on to provide capacity for hosting computing sessions; receiving, by a computing device, a configuration value of the organization indicating a probability that there will be available capacity when new computing sessions are initiated; determining, by the computing device, capacities needed to satisfy the probability at different points in time based on the historical data; and auto-scaling the host machines at one or more times according to the determined capacities.