Virtual Machine Pool Health Index for Capacity Forecasting

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In large-scale distributed computing systems, there is a lack of indication regarding the health and capacity of virtual machine pools, making it difficult for clients to determine if a pool can handle expected demand, leading to potential resource shortages and inefficiencies.

Innovation Solution

A system and method are developed to generate a pool health index value by comparing forecast demand data with forecast capacity data, providing an indicator of the relative health of virtual machine pools, which can be used by service providers and customers to ensure sufficient capacity and make informed decisions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If virtualization technologies are used to share physical computing resources among multiple customers, then resource utilization efficiency is improved, but it becomes difficult to monitor and indicate the health and capacity status of virtual machine pools

Engineering Contradiction:
Improveresource utilization efficiencyVSAvoidpool health information
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The patent introduces a pool health index as an intermediary metric that mediates between the complex internal state of virtualized computing pools and the information needs of customers. This index aggregates multiple underlying parameters (capacity utilization, job queue status, host health) into a single comprehensible indicator, solving the information loss problem while preserving resource utilization efficiency

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements feedback by continuously monitoring pool status and generating health index values that are returned to customers. This feedback loop enables customers to make informed decisions about resource allocation and job submission based on current pool conditions, addressing the information asymmetry created by virtualization

Inventive Principle:
Principle #23Feedback

2Quantity of substance

If the scale and scope of data centers are increased to serve more customers, then service capacity is improved, but the complexity of provisioning and managing physical computing resources increases

Engineering Contradiction:
Improveservice capacityVSAvoidresource management complexity
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The pool health index serves multiple functions simultaneously: it indicates current capacity status, predicts future availability, guides resource allocation decisions, and simplifies customer interactions. This multi-functionality reduces management complexity while supporting increased service capacity across large-scale data centers

Inventive Principle:
Principle #6Universality (Multi-functionality)

3Adaptability or versatility

If virtual machine pools are configured to handle diverse customer needs, then adaptability is improved, but it becomes difficult to determine if a pool can handle expected demand

Engineering Contradiction:
Improvepool configuration flexibilityVSAvoiddemand handling capability
Core Design Contradiction:
Adaptability or versatilityVSDifficulty of detecting and measuring

Solution Approach 1:

The system uses parameter changes by transforming multiple complex pool state parameters into a single health index value. This index dynamically adjusts based on underlying metrics like capacity utilization and job queue depth, making it easy to detect and measure demand handling capability regardless of pool configuration diversity

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS10042676B1Capacity pool health index
Publication Date: 2018.08.07 AMAZON TECH INC
  • US10042676B1 patent drawing
  • US10042676B1 patent drawing
  • US10042676B1 patent drawing

AI summary

A system includes one or more computer systems that include one or more processors and memory. The memory includes instructions, that, upon being executed by the one or more processors, cause the system to generate forecast demand data for a first pool of servers configured to host virtual machine instances, generate forecast capacity data that forecasts a number of slots available to run virtual machine instances for the first pool of servers, and generate a first pool health index value based on a comparison of the forecast demand data with the forecast capacity data.