October 16, 2022
Microsoft's flagship annual event, Microsoft Ignite, took place last week and as usual there were countless exciting announcements of new features and products coming to the Microsoft suite of offerings across Azure, AI, Data, Developer Experience, Security and Productivity.
The theme to Ignite this year was doing more with less, which is quite timely given the economic climate business' need to operate within today and into the future. Data and artificial intelligence shone this year in alignment with the theme, along with a big focus on productivity and security as well as several key announcements in the Azure infrastructure space. These position Microsoft as the go-to cloud and technology provider to support business' to not only survive, but thrive in today's rapidly evolving world.
Our expert engineers at Azenix have hand-picked announcements they feel are important to be aware of, will have the most impact to users, or are just outright exciting or interesting! Without further ado, here's what the team is excited about:
This service is a new way for developers to get access to custom-built VMs through a self-service platform. Dev Box allows you to create your own custom development machines with all the tooling needed for whatever projects your developers are working on. Developers can even have multiple Dev Box instances allowing for quickly switching between projects without needing to have all the tooling for each project installed onto one machine.
Dev Box can be deployed into your existing Azure infrastructure, linking to a virtual network to give the Dev Box direct access to cloud resources. This allows for more secure access to protected environments and uses the unified management, security, and compliance tooling from Microsoft Intune and Microsoft Endpoint Manager to easily control permissions and policies.
Once configured a developer just needs to log into devbox.microsoft.com and click on " + New dev box" from here the developer can create a machine based on defined Dev Box Pools. The Dev Box can be accessed either through the browser or through the Microsoft Remote Desktop app. A Dev Box Pool is the combination of a VM image and the VM specification. As the Dev Box is billed by consumption there are 2 Compute levels and 3 storage levels that can be used to create the VM spec. Each component is billed per hour so this means if the developer is not using the machine and it is stopped then only the storage component to billed.
The Microsoft DevBox can help speed up the onboarding process, improve access to cloud resources and provide a simpler way to manage developer tools across multiple projects. DevBox is in public preview now.
For more details head to the documentation here.
Azure Deployment Environments (ADE) is a service that lets your developers easily create the infrastructure for the app they are working on in a self-service way. The infrastructure is stored as Infrastructure-As-Code (IaC) templates and linked to a project. The infrastructure can be deployed either manually through a Dev Portal or the Azure CLI or can be tied to a CI/CD pipeline. As the templates are managed by the project, developers are freed up to focus on the code.
At the time of writing the IaC templates are built using ARM, there are plans to eventually bring out options for Terraform and Bicep as well. Security is managed by using a managed identity to provision the resources on behalf of authorised developers. This then grants the required permissions to the developer to create and destroy their own environments without giving broad contributor access to the entire subscription. Extensive access controls are provided by Azure RBAC and Azure AD.
Environments can be configured to auto-expire saving costs and ensuring that resources are cleaned up after they are no longer required. At the time of writing, there is no cost associated with the Azure Deployment Environments service while it is in preview. The only costs billed will be based on the services and infrastructure deployed using the service.
For more information you can read the documentation here.
Virtual machine (VM) scale sets previously required you to manually configure scale-out and scale-in rules based on metrics such as percentage CPU. Now this can be taken care of for you. With just 7 days of history your scale set can start to identify cyclical patterns where it should scale in and out. The service leverages artificial intelligence to identify patterns based on historical usage. Although you only need 7 days of data to get started, 15 days is recommended by Microsoft for more accurate results.
Predictive autoscale adheres to your existing scaling boundaries and you can configure how far in advanced the new instances will be deployed up until 1 hour before a scaling event occurs. If you want to take predictive autoscale for a test drive you can use "forecast only" mode which allows you to see your CPU forecast without triggering any scaling and then compare it to your actual workload patterns. You'll need to make sure you have rules set for scaling in as predictive autoscale currently only supports scaling out. As of today it's only capable of using the average of the "percentage CPU" metric.
You can find out more here.
Spot Priority Mix is currently in public preview and is a new feature of VM scale sets that allows you to combine standard virtual machines with spot instances to reduce costs. Spot VMs can provide up to a 90% discount compared to standard pay-as-you-go pricing, but they can be terminated if capacity is required by Microsoft with only 30 seconds notice. Being able to mix standard VMs with spot instances gives you the flexibility to have standard VMS running at all times and fill capacity with spot VMs at drastically reduced prices. You also get the confidence that your workload won't be evicted all at once should the spot instances be required to meet capacity.
You're in control with spot priority mix too, you get to choose the distribution between spot and regular VMs and configure the base number of standard VMs to run to prevent complete interruption. The service then handles scaling in and out automatically according to the parameters you set.
More details here.
Azure Reserved Instances are a staple of any cost optimization strategy, allowing organizations to commit to a certain level of usage of specific resource types. Whilst existing Azure Reserved Instances provide a certain a level of flexibility within the compute family and region via Instance Size flexibility, the new Azure Savings Plan for Compute takes this to the next level. Azure Reserved Instances require you to have a certain number of cores within the correct family each hour to maximize savings. Savings plans allow you to realize savings up to 65%, regardless of region or instance series. Savings plans can be purchased upfront, or billed monthly with either a 1-year or 3-year commitment.
From a billing perspective, usage is calculated hourly. Any spend below and up to your commitment will be billed with savings plan prices. Any spend above the commitment will be added to the invoice separately at your pay-as-you-go pricing. If your overall eligible spend is below the commitment, you will still be charged the commitment amount every hour.
Importantly, the introduction of Azure Savings Plans also means the future discontinuation of self-service exchanges and refunds for reservations. For a limited time, certain existing reservations will be available to trade in for savings plans.
Read more here.
Automanage makes maintaining and operating servers easier by automating their configuration and management. Automanage can work with Azure virtual machines and Azure Arc-enabled servers running elsewhere. It can take care of applying configuration profiles, either from best practice templates or your own custom profiles, as well as detecting and remediating drift. Even more convenient, Automanage takes care of ancillary virtual machine services such as Azure Backup, Azure Monitor and Defender for cloud. With Hotpatch and Windows Server for Azure machines Automanage can even apply patches without needing to reboot, increasing uptime and availability without sacrificing security. The best part is that it's free! You only pay for services that are deployed through the service (e.g. Azure Backup).
Find out more here.
This update makes Cosmos DB the first cloud managed database to support both relational and noSQL workloads. Microsoft has implemented PostgreSQL support by leveraging the open source Citus distribution of PostgreSQL. This enables Cosmos to have full API compatibility and have committed to staying within 2 weeks of any upstream release.Azure has also included a new PostgreSQL extension, pg_azure_storage that will allow directly copying data from an Azure Blob Storage container into a database table, removing the need for extra infrastructure such as a VM for this type of copy.I have always been a huge fan of PostgreSQL so I am very excited to have it integrated into the Cosmos DB offering and take advantage of the global availability and scalability of Cosmos.
Details are available here.
The latest version of Visual Studio 2022, v17.3 is now generally available. This version continues to build on the top-class 64-bit integrated development environment first released with .NET 6 in November last year. Microsoft continues to improve the development experience and productivity features to speed up the development lifecycle. The latest version focuses on development and tooling to help users with their .NET Multi-platform App UI (MAUI) development experience. Our team is excited about the opportunities .NET MAUI provides, as it allows us to build native mobile and desktop applications from a single codebase. With these updates to Visual Studio, the development experience is better than ever!
Download the latest version here.
We are excited about the announced partnership with Azure API Management and Postman. Postman is a tool our engineers regularly use to develop, test, and build robust APIs for our customers. The announced partnership provides integration between Azure API Management and Postman. The integration will enable our team to quickly test our Azure APIs, import and export between the services, and further accelerate the turnaround time from code to deployment.
Read more about the partnership here.
Several App Service updates were announced at Microsoft Ignite 2022. You can now speed up your WordPress development on Azure App Service with Azure Front Door that enables faster page loads, enhanced security, and increased reliability for your global apps with no configuration or additional code required.
Recently the App Service front-ends were updated to a new architecture based on kestrel and YARP, this unlocked the ability to serve gRPC-over-HTTP/2 on Azure App Service. The update in November also introduces new language support for python, Node, .NET, PHP, Wordpress and Go. Since Go and gRPC go together so well, Azure App service includes the GO language for the first time on Linux App Service, this makes it even easier to build the cloud based command line tools from within Azure.
Developers can optionally setup Application Insights for their web apps, with the new “On By Default” the telemetry is collected without the need to make any modification to the code. The “On By Default “ for Applications Insights are now enabled for .NET and Java.
App Service Environments are Premium service options for Azure App Service that delivers an enhanced configuration capability. App Service Environment v3 (ASEv3) will soon get access to powerful new SKU options in November update, with support being added for 16 vCPU/64 GB, 32 vCPU/128 GB and 64 vCPU/256 GB sizes. These new SKU address customer needs for workloads running in a dedicated tenant with higher memory requirements and more demanding requests per second (RPS) scenarios.
With the addition of new hardware also comes an interesting introduction of planned maintenance for ASEv3. The ASEv3 supports the concept of early/late upgrade preference whereby an ASEv3 is either upgraded earlier or later in App Service platform’s deployment order. With the addition of a new upgrade preference called “manual upgrade” the new planned maintenance capability notifies customers 15 days in advance of planned platform upgrades. Once notified, they can trigger the platform maintenance anytime during the next 15 days.