Hypervisor Host Deployment in Public Cloud
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In a software-defined data center, users subscribing to a public cloud under the bring-your-own-host (BYOH) model face challenges in adapting virtualization software to conform to the hardware configuration of bare-metal hosts provided by the public cloud, requiring customized hypervisor images that can be efficiently deployed and managed.
Innovation Solution
A method involving a deployment service that obtains a prototype hypervisor image, customizes it based on host and network configuration information, and stores the customized image for installation on bare-metal hosts using the public cloud's deployment API, enabling users to adapt the hypervisor to specific hardware and network settings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If users deploy their own virtualization software on bare-metal hosts in a BYOH model, then users can adapt the hypervisor to conform to specific hardware configuration requirements, but the complexity of adapting and customizing the hypervisor increases
Solution Approach 1:
The system pre-provisions customized hypervisor images that are prepared in advance with specific hardware configurations, network settings, and virtualization software adaptations. This eliminates the need for users to perform complex customization tasks manually, as the hypervisor images are already tailored to match the bare-metal host specifications before deployment.
Solution Approach 2:
A deployment service acts as an intermediary between the user's customization needs and the bare-metal hosts. This service receives hardware configuration information, applies the necessary customizations to hypervisor images, and handles the deployment process, thereby reducing the operational complexity for users while maintaining adaptability to specific hardware requirements.
2Ease of manufacture
If the public cloud provider provides bare-metal hosts with fixed hardware configuration, then hardware resource allocation is simplified, but the user's ability to adapt virtualization software to specific hardware requirements is reduced
Solution Approach 1:
The system changes the parameters of the hypervisor software by applying different customization configurations to match specific hardware requirements. The deployment service modifies hypervisor images based on the bare-metal host configuration parameters, enabling adaptation without changing the hardware itself. This allows fixed hardware provisioning while maintaining software adaptability through parameter modification.
3Manufacturing precision
If users manually customize hypervisor images for each bare-metal host, then the hypervisor can be optimized for specific hardware, but the deployment time and operational overhead increase
Solution Approach 1:
The deployment service performs preliminary customization of hypervisor images before deployment to bare-metal hosts. By pre-applying the necessary hardware-specific configurations, network settings, and virtualization adaptations in advance, the system achieves precise hardware optimization without requiring time-consuming manual customization during the deployment phase.
Solution Approach 2:
The system uses standardized hypervisor image templates that can be copied and customized efficiently. Instead of manually creating unique hypervisor configurations for each host, the deployment service copies base images and applies targeted customizations, significantly reducing deployment time while maintaining the precision required for hardware-specific optimization.
Data Source
AI summary
An example method of deploying a hypervisor to a host in a public cloud includes: obtaining, by a deployment service, a prototype hypervisor image from shared storage; obtaining, by the deployment service, configuration information for the host and a physical network of the public cloud to which the host is attached; customizing, by the deployment service, the prototype hypervisor image in response to the configuration information to generate a customized hypervisor image; storing, by the deployment service, the customized hypervisor image in the shared storage in a manner accessible by the public cloud; and invoking a deployment application programming interface (API) of the public cloud to retrieve and install the customized hypervisor image to the host.