Azure Container Services Docs – Weekly Update (2026-06-21 to 2026-06-28)

AKS – Istio Support Policy

Summary

The documentation now includes updated support timelines for Istio versions in the context of the asm-1-30 release, specifying expected availability dates for various versions. Users can refer to this information to better plan their deployments and upgrades based on the newly outlined support schedule.

AKS – Restrict Ssh Access

Summary

The documentation for restricting SSH access on AKS has been updated to reflect a more general reference to "AKS on Azure Local" instead of specifically "AKS on Azure Local 22H2." This change clarifies the applicability of the guidance, making it relevant for users working with different versions of Azure Local. Users can now better understand that the instructions apply beyond just the 22H2 version.

AKS – Use Kms Etcd Encryption

Summary

Updated guidance on handling deleted keys in KMS encryption for AKS. It now specifies the need to check for soft-deleted states and the implications of permanently deleted keys on cluster functionality.

ACR – Container Registry Dedicated Data Endpoints

Summary

The documentation now clarifies that dedicated data endpoints are a prerequisite for enabling IPv6 dual-stack endpoints in Azure Container Registry. Users can learn how to enable these dual-stack endpoints, which require the registry’s endpoint protocol to be set to `IPv4AndIPv6` after enabling dedicated data endpoints.

AKS – Use Cvm

Summary

The document now includes instructions for creating a CVM node pool using Azure Linux, along with necessary commands and notes about feature registration. It highlights that CVMs with Azure Linux are currently in preview.

AKS – Deploy Gpu Node Pool

Summary

This document provides a comprehensive guide on deploying GPU-enabled node pools in AKS on Azure Local for multi-rack deployments. It includes details on supported GPU models, VM sizes, and installation steps for the NVIDIA driver, emphasizing the importance of using the correct packages for GPU workloads.

AKS – Best Practices Gpu

Summary

The document has been significantly updated to include best practices for managing GPU-enabled node pools on AKS, emphasizing the use of AKS Automatic as the recommended default for production workloads. It outlines operational baselines, security safeguards, and SLA-backed pod readiness.

AKS – Concepts Vulnerability Management

Summary

The documentation now emphasizes that for most production workloads, AKS Automatic is the recommended default experience, which simplifies management by providing features like managed system node pools and automatic upgrades. Users can expect reduced operational overhead and enhanced security posture with AKS Automatic, while still having the option to use AKS Standard for more customized configurations. Additionally, the article details how Microsoft identifies and patches vulnerabilities across various components, including AKS container images and operating systems, ensuring a more secure environment for users.

AKS – Istio Deploy Egress

Summary

This update clarifies the scheduling behavior of Istio add-on egress gateway pods, specifying that they do not get scheduled onto the ‘gateway’ node pool. It provides guidance on using node labels for scheduling preferences.

AKS – Use Nvidia Gpu

Summary

The documentation now recommends using AKS-managed GPU node pools as the primary approach for GPU node management on AKS, with self-managed setups as an alternative. It includes a detailed comparison of the two approaches, outlining the components managed by AKS and those managed by the user.

AKS – Node Auto Provisioning Aksnodeclass

Summary

The documentation for AKSNodeClass has been revised to clarify the default OS version information for Ubuntu and the imageFamily field. It now specifies that the default OS version changes based on the Kubernetes version.

AKS – AKS Bare Metal Create Cluster Arm Template

Summary

The documentation now reflects the updated command for connecting to the cluster, changing from `az connectedk8s proxy` to `az aksarc get-credentials`. Users will need to use the new command format to successfully retrieve credentials for their AKS Arc clusters. This change enhances clarity and ensures users are following the latest practices for cluster management.

AKS – AKS Automatic Limitations

Summary

The documentation for AKS automatic limitations has been updated to remove previous restrictions regarding the Istio-based service mesh add-on, custom scraping of Prometheus metrics, log collection, and enabling ACNS observability during automatic cluster creation. Users can now utilize these features without the previously stated limitations, enhancing their ability to configure and monitor their AKS environments effectively.

AGC – How to Inference Gateway

Summary

Learn how to expose a self-hosted vLLM model server through Application Gateway for Containers inference gateway by using Gateway API Inference Extension resources.

AGC – Application Gateway for Containers Components

Summary

The documentation now clarifies the order of precedence for request timeouts in Application Gateway for Containers. Users can expect that when a request timeout is configured in multiple locations, the system will prioritize Gateway API HTTPRoute timeouts defined on the `HTTPRouteRule`, followed by `RoutePolicy` timeouts, and finally default timeout values if neither is set. This change helps users understand how timeout configurations will be applied in their setups.

AKS – Reliability Zone Resiliency Recommendations

Summary

The documentation for Azure Kubernetes Service (AKS) has been updated to provide clearer guidance on designing and validating zone resiliency, specifically for AKS Automatic and AKS Standard. Users can now learn how to make AKS cluster components zone resilient, design stateless applications for zone failure scenarios, choose appropriate storage redundancy options, and test application behavior during zonal faults. Additionally, the article emphasizes the importance of understanding application dependencies and provides detailed recommendations for scaling strategies and node management in relation to availability zones.

AKS – Auto Upgrade Cluster

Summary

The documentation now clarifies that for **AKS Automatic** clusters, upgrades are preconfigured to use the **stable** channel, automatically upgrading to the latest patch on minor version N-1 without requiring user configuration. Additionally, it emphasizes that AKS will automatically upgrade clusters in version N-3 to N-2 to maintain support compliance, and users can set planned maintenance windows for these upgrades. For **AKS Standard** clusters, users must manually select the upgrade channel that best fits their operational needs.

AKS – Scale Requirements

Summary

The documentation now includes a new section on GPU requirements, specifying that the only supported GPU model is the NVIDIA RTX Pro 6000 Blackwell with 48 GiB. Users can refer to this section for details on supported GPU-enabled VM sizes and deployment steps for compute-intensive workloads.

AGC – Overview

Summary

The documentation for Application Gateway for Containers has been updated to clarify that it is a managed ingress service for Kubernetes workloads, capable of routing HTTP, HTTPS, gRPC, and AI inference traffic to applications in Azure Kubernetes Service (AKS). Users can now configure the service using Ingress and Gateway API resources, allowing for a shared Azure-managed ingress layer while utilizing Kubernetes-native resources for traffic management. Additionally, the ALB Controller now implements version 1.5 of the Gateway API, enhancing support for various routing features and capabilities.

AKS – Ai Toolchain Operator

Summary

The documentation has been updated to clarify that only the `Windows` OS SKU is not currently supported, removing previous mentions of `AzureLinux` as unsupported. This change provides users with a clearer understanding of the OS SKU limitations when using the AI Toolchain Operator in Azure Kubernetes Service.

AKS – AKS Bare Metal Preview Limitations

Summary

The document has been updated to correct the ms.topic value from ‘conceptual’ to ‘concept-article’, improving the categorization of the content.

AGC – Quickstart Deploy Application Gateway for Containers Alb Controller Helm

Summary

The documentation has been updated to reflect the new version 1.11.1 of the Application Gateway for Containers. Users should now use the updated command syntax with `–version 1.11.1` instead of the previous version 1.10.28 when deploying the application gateway using Helm. This change ensures that users are utilizing the latest features and improvements available in the new version.

AKS – Configure AKS Scheduler

Summary

The documentation for configuring AKS scheduler profiles has been updated to correct the syntax for the `schedulerName` parameter, changing it from `default_scheduler` to `default-scheduler`. This change ensures that users implement the correct naming convention when configuring their AKS scheduler, which may prevent potential issues during setup.

AKS – Intro AKS Automatic

Summary

The documentation for AKS automatic ingress has been updated to clarify the default and optional ingress configurations. Users can now utilize the Kubernetes Gateway API via the application routing add-on for new clusters starting with AKS 1.36, and the Istio-based service mesh add-on is also highlighted as an optional feature. Additionally, the mention of Istio CNI for the Istio-based service mesh add-on has been included, providing users with more flexibility in their ingress options.

AGC – Inference Gateway

Summary

Learn how Application Gateway for Containers inference gateway supports self-hosted AI inference workloads with Gateway API Inference Extension resources.

AKS – Concepts Identity

Summary

This update enhances the documentation on identity scenarios in Azure Kubernetes Service (AKS), detailing how they apply to both AKS Automatic and AKS Standard. It emphasizes the recommended use of AKS Automatic for most production workloads, providing a production-ready platform baseline with identity-related defaults.

ACR – Container Registry Firewall Rules

Summary

The documentation now includes information about IPv6 dual-stack endpoints, explaining their prerequisites and how they affect firewall rules for Azure Container Registry.

ACR – Container Registry Ipv6 Dual Stack

Summary

This document introduces IPv6 dual-stack endpoints in Azure Container Registry, allowing clients to access the registry over both IPv4 and IPv6. It details the prerequisites, including the need for dedicated data endpoints and the Azure CLI version required for updates. The document also outlines the implications for firewall rules and interactions with other registry features.

AKS – Azure Cni Powered by Cilium

Summary

The documentation now specifies that Azure CNI Overlay powered by Cilium is the default virtual network for AKS Automatic clusters, requiring no configuration. It also details the steps for using Azure CNI Powered by Cilium in both AKS Automatic and AKS Standard clusters.

ACR – Container Registry Endpoint Reference

Summary

The endpoint reference document has been updated to include details about the new `–endpoint-protocol` command for setting the endpoint protocol to IPv4 or IPv4AndIPv6. It specifies the requirements for enabling dual-stack endpoints and links to relevant documentation.

AKS – Auto Upgrade Node Os Image

Summary

The documentation for autoupgrading node OS images in Azure Kubernetes Service (AKS) has been updated to clarify that for AKS Automatic clusters, node OS upgrades are preconfigured using the NodeImage channel. It also includes tips for adjusting maintenance windows for AKS Automatic clusters.

Fleet – Howto Use Overrides Customize Resources Placement

Summary

The documentation now clarifies that Azure Kubernetes Fleet Manager can deploy the same resource across multiple clusters while allowing for modifications to the resource configuration based on different environments (dev, test, prod). Users can now see specific examples of how to configure resource overrides, including using `clusterSelector` and `jsonPatchOverrides` in `overrideRules`, and how to apply these changes using the `kubectl apply` command. Additionally, the namespace for resource overrides has been updated to `nginx-demo` in the examples provided.

AKS – Optimize AKS Costs

Summary

This article has been updated to include new guidance on optimizing Azure Kubernetes Service (AKS) usage and costs, emphasizing AKS Automatic as the recommended default for production workloads. It covers various optimization strategies including scaling, infrastructure sizing, and GPU usage, providing a comprehensive approach to cost management.

AKS – Node Auto Provisioning

Summary

The documentation now emphasizes that AKS Automatic is the recommended default for most production workloads, as it includes node auto-provisioning (NAP) preconfigured with SLA-backed pod readiness. Users can expect reduced operational overhead for node sizing and lifecycle operations, allowing workloads to start promptly on the right-sized nodes without manual intervention. Additionally, guidance is provided to help users choose between AKS Automatic and AKS Standard based on their specific needs and scenarios.

AKS – Cluster Upgrade

Summary

The documentation now specifies that when upgrading a supported AKS cluster, users must perform upgrades sequentially by minor version number rather than major version number, which was previously stated. Additionally, the latest GA patch version for Kubernetes has been updated to 1.33.11, and users can now upgrade to Kubernetes version 1.34.7 using the provided Azure CLI command.

AKS – Workload Identity Overview

Summary

The documentation now clarifies that on **AKS Automatic** clusters, workload identity and the OIDC cluster issuer are preconfigured by default, eliminating the need for cluster-level setup before using workload identity in pods. Users can directly configure their applications to utilize this identity, whereas on **AKS Standard** clusters, they must enable and configure workload identity separately. Additionally, the section on identity bindings has been updated to emphasize its role in supporting large-scale AKS environments.

AKS – AKS Bare Metal System Requirements

Summary

This document has been updated to reflect the latest system requirements for AKS on bare metal, including compatibility and configuration details.

AGC – Service Mesh Integration

Summary

The documentation has been updated to reflect the new version 1.11.1 of the Application Gateway for Containers, replacing the previous version 1.10.28. Users can now access features and improvements introduced in version 1.11.1, enhancing their experience with service mesh integration.

AKS – Use Node Public IPs

Summary

The documentation now includes a new section on using dual-stack public IP prefixes in Azure Kubernetes Service (AKS), allowing users to assign both IPv4 and IPv6 public IP addresses to each node in a node pool. Users must register the `NodePublicIPv6PrefixPreview` feature flag and meet specific prerequisites, such as configuring their cluster as dual-stack and using Azure CNI with overlay networking. Additionally, users can create new clusters or add node pools with dual-stack public IP prefixes by utilizing the `–node-public-ip-prefix-ids` parameter.

AKS – Istio Deploy Ingress

Summary

The documentation now clarifies how to schedule ingress gateway pods onto specific nodes using the `azureservicemesh/istio.replica.preferred` node label or AKS system nodes. It specifies that for the asm-1-30 release, the weighted preferences for node affinities have changed, with a preference of `100` for the `azureservicemesh/istio.replica.preferred: true` nodes and `50` for AKS system nodes. Additionally, it notes that for earlier revisions, these weights were reversed, providing users with important context for configuration changes.

AKS – AKS Bare Metal Create Cluster Bicep

Summary

The command for connecting to the cluster has been updated to use ‘az aksarc get-credentials’, reflecting changes in the Azure CLI commands for AKS Arc.

AKS – AKS Bare Metal Create Cluster Portal

Summary

The documentation for the control plane IP configuration has been updated to clarify that if the control plane IP field is left blank, it will default to the host machine’s IP address. Users are now informed that if they specify a custom control plane IP while using DHCP, they must reserve that IP to ensure it remains permanently assigned; otherwise, changes to the control plane IP could render the Kubernetes cluster unreachable and necessitate redeployment.

AGC – How to Cors Gateway API

Summary

Learn how to configure Cross-Origin Resource Sharing (CORS) in Gateway API for Application Gateway for Containers.

AGC – Service Mesh Helm Chart

Summary

The documentation for the alb-controller-servicemesh-extension has been updated to clarify the parameters related to image configuration and istio extension settings. Users can now find detailed descriptions for the image parameters, including the image name, pull policy, registry, and tag, as well as the configuration for the istio extension, which includes resource limits and requests. Additionally, the log level and provider information have been explicitly defined, ensuring users have a clearer understanding of the configuration options available.

AKS – Tutorial Kubernetes Prepare App

Summary

The documentation has been updated to reflect the use of RabbitMQ version 4.3.2-management-alpine in the AKS tutorials, replacing the previous version 3.13.2-management-alpine. Additionally, unnecessary configurations related to volumes and environment variables have been removed, streamlining the setup process for users. This change allows users to work with a more current version of RabbitMQ without the clutter of outdated configurations.

AGC – API Specification Kubernetes

Summary

The documentation now clarifies that if an endpoint does not report load metrics for a duration of 3 minutes, the system will stop using those metrics to influence load balancing decisions. Additionally, it specifies that users can utilize at most one filter on a rule for forwarding, and that values might be added to the enum in future implementations. These changes enhance the understanding of load balancing behavior and rule configuration for users.

AKS – Tutorial Kubernetes Paas Services

Summary

The documentation has been updated to remove the `rabbitmq` StatefulSet, ConfigMap, and Service sections from the AKS tutorials, streamlining the content. Additionally, the `order-service` Deployment section has been revised to reflect these changes, which may impact how users configure their applications in Azure Kubernetes Service. Users will need to adjust their deployments accordingly, as the previous configurations for RabbitMQ are no longer included.

AGC – Cross Origin Resource Sharing

Summary

Learn how Cross-Origin Resource Sharing (CORS) works with Application Gateway for Containers.

AKS – Concepts Ai Ml Language Models

Summary

The documentation now emphasizes that for most production AI and machine learning workloads on Azure Kubernetes Service (AKS), AKS Automatic is the recommended default, providing production-ready configurations for scaling, security, and operations. Users can leverage AKS Automatic to reduce operational overhead and focus more on model onboarding and evaluation rather than cluster management. Additionally, the content has been updated to clarify the distinctions between small and large language models, including their parameter counts and use cases.

AGC – Alb Controller Release Notes

Summary

The documentation now includes updates for version 1.11.1 of the ALB Controller, which implements Gateway API v1.5.1 and adds support for the Gateway API Inference Extension, Arm64 nodes, CORS, HTTPRoute timeouts, and HTTP redirect 307/308. Users can also benefit from fixes related to helm subcharts and logging enums, enhancing overall functionality and performance.

AKS – Blue Green Node Pool Upgrade

Summary

This document has been updated to include new information about undrainable node behavior and maxBlockedNodes settings. It provides guidance on managing node upgrades effectively, including links to relevant configurations.

AKS – Tutorial Kubernetes Deploy Application

Summary

The documentation for deploying applications in Azure Kubernetes Service (AKS) has been updated to reflect the removal of unnecessary configurations and the update of the RabbitMQ version. Users can now follow the revised tutorial to deploy RabbitMQ with a streamlined configuration, enhancing the deployment process and ensuring compatibility with the latest version.

AGC – Alb Controller Helm Chart

Summary

The documentation now clarifies that users can configure additional parameters during the installation of the ALB Controller, specifically `aiGateway` for enabling AI workload support and `installInferenceExtensionCRDs` for managing Inference Extension CRDs. Additionally, the `cloudEnvironment` parameter allows users to specify the Azure cloud environment, enhancing deployment flexibility. The examples and descriptions have been updated for clarity, ensuring users understand how to utilize these new features effectively.

AKS – Howto Enable Windows Node Pools

Summary

This update includes a note about GMSA authentication not being supported on Windows node pools, along with a new date for the document. It enhances clarity on the limitations of Windows node pools in AKS.

AKS – Howto Create Windows Node Pools

Summary

The documentation now includes a note indicating that Group Managed Service Accounts (gMSA) authentication is not supported on Windows node pools, which is crucial for users planning to implement gMSA in their AKS environments. Additionally, the instructions for creating a sample YAML file have been updated to emphasize the file name formatting.

Fleet – Faq

Summary

New content has been added to explain the transition of the Fleet Manager hub cluster from a ‘Failed’ to ‘Running’ state, detailing the automatic reconciliation process and its implications for cost.

AKS – Azure Linux Retirement

Summary

The Azure Kubernetes Service (AKS) will no longer support Azure Linux 2.0 starting November 30, 2025, with security updates ceasing. Users are advised to migrate to a supported Azure Linux version by upgrading their node pools or migrating to AzureLinux3. This change is crucial for maintaining security and operational capabilities.

AKS – Howto Upgrade Windows Os

Summary

The documentation now includes a note indicating that Group Managed Service Accounts (gMSA) authentication is not supported on Windows node pools. This update clarifies limitations for users configuring authentication in their AKS environments, ensuring they are aware of this restriction when planning their deployments.

Fleet – Concepts Fleet Arc Integration

Summary

The documentation now includes specific resource requirements for deploying the Fleet Manager Arc extension on Arc-enabled Kubernetes clusters, which require at least 210 MB of memory and 2% of one CPU core, along with the reservation of 3 pods for the extension agents. Additionally, the section on progressive rollouts has been updated to clarify that it utilizes staged update runs, allowing application owners to manage rollouts more effectively. Users are also advised to review considerations before adding Arc-enabled clusters to Fleet Manager, ensuring they understand the implications of this integration.

AKS – AKS Customer Proxy

Summary

Learn how to bring your own HTTPS proxy to provide internet connectivity for AKS on Azure Local multi-rack clusters.

AKS – Gmsa Windows Node Pools Not Supported

Summary

This new section clarifies that gMSA authentication is not supported for Windows node pools on Azure Local 23H2 and 24H2. It explains the limitations and provides guidance for customers needing gMSA for Active Directory access, suggesting alternatives.

AKS – Upgrade Options

Summary

This article has been significantly updated to include scenario-based recommendations for upgrade challenges and emphasizes AKS Automatic as the recommended default for production workloads. It also provides a technical foundation for AKS cluster upgrades.

AGC – Server Sent Events

Summary

The documentation now recommends using Gateway API primitives by defining a `timeouts.request` value of `0s` on the `HTTPRouteRule`, replacing the previous guidance to define a `RoutePolicy` resource with a `routeTimeout` value of `0s`. Additionally, a note clarifies that if a Gateway API HTTPRoute timeout is defined, it takes precedence over the `RoutePolicy` timeout. Users should also be aware that server-sent events are not supported using the Ingress API.

AKS – Istio Cni

Summary

The documentation now specifies that starting with revision `asm-1-30`, Istio CNI is the default mechanism for traffic redirection in new mesh installations, eliminating the need for privileged init containers. Users of revisions `asm-1-25` through `asm-1-29` must explicitly enable Istio CNI using the `–proxy-redirection-mechanism` parameter, while earlier revisions do not support Istio CNI. This change enhances security by reducing the need for privileged capabilities in application pods.