Dynamic Virtual CPU Model Updates Without VM Shutdown

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional virtualization environments require shutting down virtual machines to update or modify virtual central processing unit (vCPU) models, resulting in undesirable downtime.

Innovation Solution

A system and method for dynamically updating vCPU models during runtime by using a virtualization manager to send requests to hypervisors for microcode updates, which allow adding, updating, or removing CPU model features without shutting down the virtual machine, by managing a hypervisor data store with active and inactive flags for microcode updates and corresponding CPU models.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If virtual machines are shut down to update vCPU models, then microcode updates can be applied, but downtime increases

Engineering Contradiction:

Improvemicrocode update applicationVSAvoiddowntime

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system dynamically switches between active and inactive CPU model versions during runtime. The hypervisor maintains multiple CPU model definitions with version identifiers and dynamically transitions between them without requiring virtual machine shutdown, enabling continuous operation while applying microcode updates.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The system prepares inactive CPU model versions in advance before they are needed. Microcode updates are staged in inactive model definitions that are pre-configured but not yet active, allowing the system to switch to the updated model seamlessly when ready, without interruption to the running virtual machine.

Inventive Principle:

Principle #10Preliminary action

2Reliability

If CPU model features are modified, then security and stability are enhanced, but system complexity increases

Engineering Contradiction:

Improvesecurity and stabilityVSAvoidhypervisor data store management

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The hypervisor data store is segmented into multiple CPU model definitions, each with unique version identifiers and feature sets. This segmentation allows independent management of different CPU model versions, enabling the hypervisor to maintain and switch between multiple configurations without conflating their management complexity.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system implements version identifier tracking and comparison mechanisms that provide feedback on which CPU model version is currently active and what updates are available. This feedback system automates the decision-making process for model switching, reducing the manual complexity of managing multiple CPU model versions.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS10776150B2Dynamically updating virtual central processing units

Publication Date: 2020.09.15 RED HAT INC

AI summary

A method includes receiving, by a hypervisor running on a host computer system, a request pertaining to a microcode update from a guest operating system of a virtual machine running on the host computer system. The method also includes identifying, by a hypervisor, a central processing unit (CPU) model including one or more features associated with the microcode update. The method also includes emulating, by the hypervisor, the CPU model on a virtual central processing unit (vCPU) of the virtual machine to provide access to the one or more features of the CPU model to the guest operating system of the virtual machine.