In a recent blog post, Microsoft announced support for Azure IoT Edge on virtual machines. This new capability expands Microsoft’s support for virtualization platforms, including VMware vSphere and Hyper-V. Using these virtualization platforms, customers can choose to run Azure IoT Edge on Windows, a variety of Linux distributions, or containers.
Azure IoT Edge has been generally available (GA) since June 2018 and provides the ability for IoT devices to respond to events, without the need for real-time access to the internet. Some of the use cases that Microsoft targets include running local machine learning models for anomaly detection and predictive maintenance scenarios. More specifically, Microsoft has demonstrated a drone recognizing a crack in a gas pipe and predicting equipment failures before they actually happen.
Microsoft has a strategy of supporting a wide array of devices and platforms for its Azure IoT Edge offering. Chipalo Street, principal program manager at Microsoft, explains why this is important:
Azure’s intelligent edge portfolio is designed to run on a breadth of hardware to match our customers’ scenarios. This includes everything from microcontroller units (MCUs) running Azure Sphere to a fully consistent experience that is both cloud and edge, powered by Azure Stack. Azure IoT Edge already supports a variety of Linux and Windows operating systems as well as a spectrum of hardware from devices smaller than a Raspberry Pi to servers.
Supporting IoT Edge in VMware vSphere offers even more customer choice for those who want to run AI on infrastructure they already own. We see customers building hybrid cloud and edge solutions in virtually every industry, and the hardware they choose for each is fit for purpose.
As part of this announcement, Microsoft has entered into a partnership with VMware that will enable a simplified deployment experience for VMware vSphere. In a VMware blog post, Chris Wolf and Daniel Beveridge, from the office of the CTO, expressed why this partnership is important:
We see a growing need across our customer base to decouple cloud services from cloud data centers to meet many of the [networking, security and scale] requirements. In addition, oftentimes a legacy portion of the application stack can limit an application’s ability to integrate with a variety of cloud services.
Using the Microsoft-VMware solution brings Azure services like Azure Stream Analytics, Azure Functions and Azure Cognitive Services to the edge of an on-premises data center. The supported configuration includes Azure IoT Edge running in an Ubuntu Server, which can be integrated into an existing operational management stack where VMware tools can be used to operate the underlying platform including vCenter, vRealize Operations and NSX. Beyond the management opportunities that emerge with vSphere 6.7 (and later), passthrough support for Trusted Platform Module (TPM) leverages hardware root of trust and provides advanced security.
Azure IoT Edge’s runtime has been open-sourced and is available on GitHub.