Many new Azure releases last week at Build
There were many new releases and features added to Azure last week announced at the Build conference. Here are a few releases that are relevant to education:
Azure App Service – GA
The App Service release is composed of Web Apps, Mobile Apps, API Apps, Function Apps, and Logic Apps all residing on a hyper-scale platform as a service so there are no Virtual Machines to manage, patch, etc. This is a Platform as a Service offering for campuses to host their applications.
Web applications in App Service. For more on web apps see here.
There are over 83 out of the box Web applications you can deploy to Azure App Service from Azure Marketplace. See here:
Ready to go prebuilt Web App applications
Creating a mobile backend for a new or existing mobile app. For more on mobile app see here
Creating an API app for new or existing app. For more on API app see here
Logic App design for your logic apps workflow. For more on Logic apps see here.
Autoscale your App Service application based on various criteria/load
DevTest Labs – close to GA
This is a very popular feature in education since it allows you to quickly create lab environments for classrooms, student labs, research labs, etc. Some of the main benefits in education of this scenario are you can restrict VM type, cost threshold. See here for more info on DevTest labs.
See costs for lab spend and estimated cost trends
Pick the VM max sizes for the lab to control costs
Shutdown the lab VMs everyday to control costs.
Pick a VM formula to pattern the lab VMs after.
Azure Service Fabric – GA
Service Fabric is a feature-rich microservices application platform with lifecycle management, stateless and stateful services, and performance at scale. This can be used at campuses looking for a microservices platform to host their campus microservice applications.
Watch a great video on Service Fabric from Mark Russinovich here.
Setting up a Service Fabric Cluster
Example of microservices running in a Service Fabric cluster
Azure Functions – Preview
Event driven compute on demand which runs on top of App Service where you can implement code triggered events which occur anywhere in Azure (web, mobile, IoT, etc.), 3rd party or even on prem. You can go to Functions.Azure.com to get started.
Create New or leverage prebuilt function apps
Debug and modify function app real time
Template Export – GA
This is a great feature on how to export Azure templates into JSON format if you built your Azure resources in the GUI but want to recreate them using JSON scripts. See my blog post here on the feature.
Export Resource Group to JSON template and it precreates PowerShell to leverage with the JSON template exported.
Deploy from an existing exported JSON template to another Azure Resource Group which is great for Dev/Test replicas for example.
Virtual Machine Scale Sets – GA
Deploy and manage identical virtual machines with either Windows or Linux VMs as a set with full autoscale on demand. This is a great feature for university/research big compute scenarios, big data or containerized workloads.
See more on Virtual Machine Scale sets here.
Cortana Intelligence Suite
We renamed the Cortana Analytics Suite to the Cortana Intelligence Suite (CIS) which is comprised of information management, Big data stores, machine learning and analytics, and intelligence. See more on CIS here. This suite can be used in multiple campus scenarios such as Campus IoT, Campus data needs, student success analytics, campus kiosks, advanced cognitive scenarios, and many other scenarios.
Cortana Intelligence Suite services visual
Some new cognitive services you can leverage for vision, speech, emotion, face, text, etc. within your applications. See more at the gallery here.
R Server in HDInsight – GA
This is a new feature with HDInsight Premium where we include R server when creating your HDInsight clusters in Azure. This allows for campus/research data scientists, R programmers and statisticians with real time on demand access to analytics on HDInsight.
RStudio R Server with HDInsight cluster in Azure running a validation script
Read more on setting up R Server in HDInsight here.
Azure Batch on Linux – Preview
Azure Batch allows you to run large scale jobs in parallel. This is very popular feature for large research projects like engineering modeling, or financial risk models. Last week the Azure team launched Azure Batch on Linux in preview. This provides you a ‘scheduling as a service’ feature for Linux applications and workflows. You can also combine this with Docker to combine containers with Batch. See post here for more information on Azure Batch on Linux. For more on Azure Batch see here.
32 node Linux VMs running a NAMD simulation with Azure Batch on Linux