Skip to content

Latest commit

 

History

History
42 lines (30 loc) · 2.89 KB

nested-virtualization.md

File metadata and controls

42 lines (30 loc) · 2.89 KB
title description author manager ms.author ms.date ms.topic ms.service services monikerRange
Nested virtualization for Azure IoT Edge for Linux on Windows | Microsoft Docs
Learn about how to navigate nested virtualization in Azure IoT Edge for Linux on Windows.
fcabrera
kgremban
fcabrera
2/24/2021
conceptual
iot-edge
iot-edge
=iotedge-2018-06

Nested virtualization for Azure IoT Edge for Linux on Windows

There are two forms of nested virtualization compatible with Azure IoT Edge for Linux on Windows. Users can choose to deploy through a local VM or Azure VM. This article will provide users clarity on which option is best for their scenario and provide insight into configuration requirements.

Note

Ensure to enable one networking option for nested virtualization. Failing to do so will result in EFLOW installation errors.

Deployment on local VM

This is the baseline approach for any Windows VM that hosts Azure IoT Edge for Linux on Windows. For this case, nested virtualization needs to be enabled before starting the deployment. Read Run Hyper-V in a Virtual Machine with Nested Virtualization for more information on how to configure this scenario.

If you are using Windows Server, make sure you install the Hyper-V role.

Deployment on Azure VMs

Azure IoT Edge for Linux on Windows is not compatible on an Azure VM running the Server SKU unless a script is executed that brings up a default switch. For more information on how to bring up a default switch, see the Windows Server section below.

Note

Any Azure VMs that is supposed to host EFLOW must be a VM that supports nested virtualization

Windows Server

For Windows Server users, note that Azure IoT Edge for Linux on Windows does not automatically support the default switch. Before you can deploy IoT Edge for Linux on Windows you must set up an internal switch on the server.

Our deployment functionality does not create the default switch automatically because that requires IP configuration for the internal switch, a NAT configuration, and installing and configuring a DHCP server. Our deployment functionality states that it does not fiddle around with these settings in order to not affect network configurations on productive deployments.