Welcome to Microsoft Build 2023. The evolution of the Kubernetes platform has not stopped over the past year, and the Azure Kubernetes team has continued to drive innovation and features to help our customers keep up to date and make the adoption and use of Kubernetes on AKS as seamless as possible.
The value Kubernetes provides for you to ship faster, react to customers’ demands quicker whilst at the same time giving you an Enterprise platform that keeps up with those demands is at the centre of how we drive AKS forward for our customers.
Simplifying your developer experience
Getting your engineering teams up and running to start their development process would usually take process, operational and a bit of development glue to deploy your first application. Automated deployments greatly reduce the friction to onboard your teams by automatically creating GitHub Actions to build, containerize and deploy your applications to your AKS cluster. Your teams will be able to iterate quickly and focus on what matters, delivering value through code. We have also made additions to the VSCode developer extension to fully support GitHub actions within your development environment.
Automated deployments for Azure Kubernetes Service simplify the process of setting up a GitHub Action and creating a pipeline for your code releases. You can point automated deployments at your GitHub repository that contains a Dockerfile and Kubernetes manifests and we’ll create the pipeline to continuously build and release your application. Today, we’re announcing that you can now also create automated deployment pipelines even if your repository doesn’t have a Dockerfile or Kubernetes manifests. Powered by Draft. Automated deployments will now generate an appropriate Dockerfile for your application and get your apps deployed to the cluster.
The application routing add-on enables you to expose your applications using a managed, in-cluster, ingress controller that integrates with Azure Key Vault and Azure DNS. We’re announcing that you can now configure ingresses for your applications through the Azure portal, without writing any YAML files. You can pick a domain to manage through an Azure DNS zone and choose an SSL/TLS certificate from an Azure Key Vault. The add-on configures the integration and the required roles and permissions.
Operating Kubernetes at scale
AKS Fleet Manager continues to grow in functionality to address the challenges our customers have around managing multiple AKS clusters. Co-ordinating updates to your AKS infrastructure are tied to the stage in the application lifecycle (dev/test/prod), and for customers with a large footprint, the operational overhead can be immense to manage and execute.
With the introduction of update groups, stages, and runs, you are in full control on how your update your clusters and under what conditions to move to the next deployment stage.
Windows workloads on AKS continue to increase as our customers look to develop their cloud native offerings on AKS for both Operating Systems. We are announcing support for Windows node configuration, allowing you to modify kubelet settings to tune how that infrastructure works. We have also announced the General Availability of Gen 2 VM support for Windows nodes, opening the advance capabilities of those SKUs.
Microsoft’s Mariner Linux distribution, specifically designed as a Container OS for workloads has now reached General Availability and gets a new name in Azure Linux. Being able to provide support for the infrastructure, Operating System and Kubernetes platform helps Microsoft to take ownership of the software supply chain directly to manage the things you shouldn’t have to worry about.
Secure by default
Azure AD Workload Identity is now Generally available. This is an evolution of how your workloads securely access Azure resources through Kubernetes native capabilities to federate to external identity provides. Your workloads on both Linux & Windows can take full advantage of workload identity, and your applications can utilise the Azure SDK’s as well as our MSAL libraries to integrate with Azure services securely.
Connectivity and Observability
It’s always felt like Cloud Native apps and Prometheus metrics gathering goes together since the beginning of time. Observability of distributed systems is essential to monitor and maintain your service levels and to keep track of improvements in your application over time. We are delighted to announce the General Availability of managed Prometheus and Grafana; this greatly reduces the work needed to spin up, managed and maintain these critical components, letting you concentrate on your Apps.
One of the great things about Azure CNI as your network provider is that you get all the goodness that comes from the native Azure network infrastructure. Especially for Cloud Native workloads, where scaling to respond to customers’ demands, or events, it’s difficult to keep up with how much IP address space is needed for your infrastructure and Pods. With the GA of Overlay Network, you no longer need to pull IP addresses for your workloads from a scarce pool, those IP addresses can be allocated from an overlay network. So, whether you only need 3 IP addresses or 300 for your pods, you only need to worry about one address per AKS node.
Kubernetes Apps GA
And finally, we have announced the General Availability of Kubernetes Apps; a partner ecosystem specifically built for ISV partners to easily deploy their applications to AKS clusters. As well as a marketplace of offerings, we also take care of the deployment, billing and lifecycle of the apps deployed. This greatly simplifies the ease with which our joint customers can deploy cloud native applications and get support on AKS.