This article provides answers to some of the most common questions about Azure Kubernetes Service (AKS).
AKS provides SLA guarantees in the Standard pricing tier with the Uptime SLA feature.
Platform support is a reduced support plan for unsupported "N-3" version clusters. Platform support only includes Azure infrastructure support.
For more information, see the platform support policy.
Yes, AKS initiates auto-upgrades for unsupported clusters. When a cluster in an n-3 version (where n is the latest supported AKS GA minor version) is about to drop to n-4, AKS automatically upgrades the cluster to n-2 to remain in an AKS support policy.
For more information, see Supported Kubernetes versions, Planned maintenance windows, and Automatic upgrades.
Yes, AKS supports Windows Server containers. For more information, see the Windows Server on AKS FAQ.
AKS agent nodes are billed as standard Azure virtual machines. If you purchased Azure reservations for the VM size that you're using in AKS, those discounts are automatically applied.
No, moving your AKS cluster between tenants is currently unsupported.
No, moving your AKS cluster between subscriptions is currently unsupported.
No, moving or renaming your AKS cluster and its associated resources isn't supported.
No, you can't restore your cluster after deleting it. When you delete your cluster, the node resource group and all its resources are also deleted.
If you want to keep any of your resources, move them to another resource group before deleting your cluster. If you want to protect against accidental deletes, you can lock the AKS managed resource group hosting your cluster resources using Node resource group lockdown.
You can completely stop a running AKS cluster or scale or autoscale all or specific User
node pools to zero.
You can't directly scale system node pools to zero.
No, scale operations using the Virtual Machine Scale Set APIs aren't supported. You can use the AKS APIs (az aks scale
).
No, scale operations using the Virtual Machine Scale Set APIs aren't supported. You can use the AKS API to scale non-system node pools to zero or stop your cluster instead.
No, this isn't a supported configuration. Stop your cluster instead.
AKS builds upon many Azure infrastructure resources, including Virtual Machine Scale Sets, virtual networks, and managed disks. These integrations enable you to apply many of the core capabilities of the Azure platform within the managed Kubernetes environment provided by AKS. For example, most Azure virtual machine types can be used directly with AKS and Azure Reservations can be used to receive discounts on those resources automatically.
To enable this architecture, each AKS deployment spans two resource groups:
- You create the first resource group. This group contains only the Kubernetes service resource. The AKS resource provider automatically creates the second resource group during deployment. An example of the second resource group is MC_myResourceGroup_myAKSCluster_eastus. For information on how to specify the name of this second resource group, see the next section.
- The second resource group, known as the node resource group, contains all of the infrastructure resources associated with the cluster. These resources include the Kubernetes node VMs, virtual networking, and storage. By default, the node resource group has a name like MC_myResourceGroup_myAKSCluster_eastus. AKS automatically deletes the node resource group whenever you delete the cluster. You should only use this resource group for resources that share the cluster's lifecycle.
Note
Modifying any resource under the node resource group in the AKS cluster is an unsupported action and will cause cluster operation failures. You can prevent changes from being made to the node resource group by blocking users from modifying resources managed by the AKS cluster.
By default, AKS names the node resource group MC_resourcegroupname_clustername_location, but you can provide your own name.
To specify your own resource group name, install the aks-preview Azure CLI extension version 0.3.2 or later. When you create an AKS cluster using the [az aks create
][az-aks-create] command, use the --node-resource-group
parameter and specify a name for the resource group. If you use an Azure Resource Manager template to deploy an AKS cluster, you can define the resource group name using the nodeResourceGroup property.
- The Azure resource provider automatically creates the secondary resource group.
- You can specify a custom resource group name only when you're creating the cluster.
As you work with the node resource group, keep in mind that you can't:
- Specify an existing resource group for the node resource group.
- Specify a different subscription for the node resource group.
- Change the node resource group name after the cluster has been created.
- Specify names for the managed resources within the node resource group.
- Modify or delete Azure-created tags of managed resources within the node resource group.
You might get unexpected scaling and upgrading errors if you modify or delete Azure-created tags and other resource properties in the node resource group. AKS allows you to create and modify custom tags created by end users, and you can add those tags when creating a node pool. You might want to create or modify custom tags, for example, to assign a business unit or cost center. Another option is to create Azure Policies with a scope on the managed resource group.
Azure-created tags are created for their respective Azure Services and should always be allowed. For AKS, there are the aks-managed
and k8s-azure
tags. Modifying any Azure-created tags on resources under the node resource group in the AKS cluster is an unsupported action, which breaks the service-level objective (SLO).
Note
In the past, the tag name "Owner" was reserved for AKS to manage the public IP that is assigned on front end IP of the load balancer. Now, services follow use the aks-managed
prefix. For legacy resources, don't use Azure policies to apply the "Owner" tag name. Otherwise, all resources on your AKS cluster deployment and update operations will break. This doesn't apply to newly created resources.
For a complete list of available regions, see AKS regions and availability.
No, AKS clusters are regional resources and can't span regions. See best practices for business continuity and disaster recovery for guidance on how to create an architecture that includes multiple regions.
Yes, you can deploy an AKS cluster across one or more availability zones in regions that support them.
Yes, you can use different virtual machine sizes in your AKS cluster by creating multiple node pools.
AKS doesn't set a limit on the container image size. However, it's important to understand that the larger the image, the higher the memory demand. A larger size could potentially exceed resource limits or the overall available memory of worker nodes. By default, memory for VM size Standard_DS2_v2 for an AKS cluster is set to 7 GiB.
When a container image is excessively large, as in the Terabyte (TBs) range, kubelet might not be able to pull it from your container registry to a node due to lack of disk space.
For Windows Server nodes, Windows Update doesn't automatically run and apply the latest updates. On a regular schedule around the Windows Update release cycle and your own validation process, you should perform an upgrade on the cluster and the Windows Server node pool(s) in your AKS cluster. This upgrade process creates nodes that run the latest Windows Server image and patches, then removes the older nodes. For more information on this process, see Upgrade a node pool in AKS.
The following images have functional requirements to "Run as Root" and exceptions must be filed for any policies:
- mcr.microsoft.com/oss/kubernetes/coredns
- mcr.microsoft.com/azuremonitor/containerinsights/ciprod
- mcr.microsoft.com/oss/calico/node
- mcr.microsoft.com/oss/kubernetes-csi/azuredisk-csi
Yes, there are two options for limiting access to the API server:
- Use API Server Authorized IP Ranges if you want to maintain a public endpoint for the API server but restrict access to a set of trusted IP ranges.
- Use a private cluster if you want to limit the API server to only be accessible from within your virtual network.
AKS patches CVEs that have a "vendor fix" every week. CVEs without a fix are waiting on a "vendor fix" before they can be remediated. The AKS images are automatically updated inside of 30 days. We recommend you apply an updated Node Image on a regular cadence to ensure that latest patched images and OS patches are all applied and current. You can do this using one of the following methods:
- Manually, through the Azure portal or the Azure CLI.
- By upgrading your AKS cluster. The cluster upgrades cordon and drain nodes automatically and then bring a new node online with the latest Ubuntu image and a new patch version or a minor Kubernetes version. For more information, see Upgrade an AKS cluster.
- By using node image upgrade.
Microsoft provides guidance for other actions you can take to secure your workloads through services like Microsoft Defender for Containers. The following security threat is related to AKS and Kubernetes that you should be aware of:
- New large-scale campaign targets Kubeflow (June 8, 2021).
No, all data is stored in the cluster's region.
Traditionally if your pod is running as a nonroot user (which you should), you must specify a fsGroup
inside the pod's security context so the volume can be readable and writable by the Pod. This requirement is covered in more detail here.
A side effect of setting fsGroup
is that each time a volume is mounted, Kubernetes must recursively chown()
and chmod()
all the files and directories inside the volume (with a few exceptions noted below). This scenario happens even if group ownership of the volume already matches the requested fsGroup
. It can be expensive for larger volumes with lots of small files, which can cause pod startup to take a long time. This scenario has been a known problem before v1.20, and the workaround is setting the Pod run as root:
apiVersion: v1
kind: Pod
metadata:
name: security-context-demo
spec:
securityContext:
runAsUser: 0
fsGroup: 0
The issue has been resolved with Kubernetes version 1.20. For more information, see Kubernetes 1.20: Granular Control of Volume Permission Changes.
AKS uses a secure tunnel communication to allow the api-server and individual node kubelets to communicate even on separate virtual networks. The tunnel is secured through mTLS encryption. The current main tunnel that is used by AKS is Konnectivity, previously known as apiserver-network-proxy. Verify all network rules follow the Azure required network rules and FQDNs.
Yes, you can add the annotation kubernetes.azure.com/set-kube-service-host-fqdn
to pods to set the KUBERNETES_SERVICE_HOST
variable to the domain name of the API server instead of the in-cluster service IP. This is useful in cases where your cluster egress is done via a layer 7 firewall, such as when using Azure Firewall with Application Rules.
AKS doesn't apply Network Security Groups (NSGs) to its subnet and doesn't modify any of the NSGs associated with that subnet. AKS only modifies the network interfaces NSGs settings. If you're using CNI, you also must ensure the security rules in the NSGs allow traffic between the node and pod CIDR ranges. If you're using kubenet, you must also ensure the security rules in the NSGs allow traffic between the node and pod CIDR. For more information, see Network security groups.
AKS nodes run the "chrony" service, which pulls time from the localhost. Containers running on pods get the time from the AKS nodes. Applications launched inside a container use time from the container of the pod.
No, AKS is a managed service, and manipulation of the IaaS resources isn't supported. To install custom components, use the Kubernetes APIs and mechanisms. For example, use DaemonSets to install required components.
What Kubernetes admission controllers does AKS support? Can admission controllers be added or removed?
AKS supports the following admission controllers:
- NamespaceLifecycle
- LimitRanger
- ServiceAccount
- DefaultIngressClass
- DefaultStorageClass
- DefaultTolerationSeconds
- MutatingAdmissionWebhook
- ValidatingAdmissionWebhook
- ResourceQuota
- PodNodeSelector
- PodTolerationRestriction
- ExtendedResourceToleration
Currently, you can't modify the list of admission controllers in AKS.
Yes, you can use admission controller webhooks on AKS. It's recommended you exclude internal AKS namespaces, which are marked with the control-plane label. For example:
namespaceSelector:
matchExpressions:
- key: control-plane
operator: DoesNotExist
AKS firewalls the API server egress so your admission controller webhooks need to be accessible from within the cluster.
To protect the stability of the system and prevent custom admission controllers from impacting internal services in the kube-system, namespace AKS has an Admissions Enforcer, which automatically excludes kube-system and AKS internal namespaces. This service ensures the custom admission controllers don't affect the services running in kube-system.
If you have a critical use case for deploying something on kube-system (not recommended) in support of your custom admission webhook, you may add the following label or annotation so that Admissions Enforcer ignores it.
Label: "admissions.enforcer/disabled": "true"
or Annotation: "admissions.enforcer/disabled": true
Azure Key Vault Provider for Secrets Store CSI Driver provides native integration of Azure Key Vault into AKS.
FIPS-enabled nodes are now supported on Linux-based node pools. For more information, see Add a FIPS-enabled node pool.
Any patch, including a security patch, is automatically applied to the AKS cluster. Anything bigger than a patch, like major or minor version changes (which can have breaking changes to your deployed objects), is updated when you update your cluster if a new release is available. You can find when a new release is available by visiting the AKS release notes.
What is the purpose of the AKS Linux Extension I see installed on my Linux Virtual Machine Scale Sets instances?
The AKS Linux Extension is an Azure VM extension that installs and configures monitoring tools on Kubernetes worker nodes. The extension is installed on all new and existing Linux nodes. It configures the following monitoring tools:
- Node-exporter: Collects hardware telemetry from the virtual machine and makes it available using a metrics endpoint. Then, a monitoring tool, such as Prometheus, is able to scrap these metrics.
- Node-problem-detector: Aims to make various node problems visible to upstream layers in the cluster management stack. It's a systemd unit that runs on each node, detects node problems, and reports them to the cluster's API server using Events and NodeConditions.
- ig: An eBPF-powered open-source framework for debugging and observing Linux and Kubernetes systems. It provides a set of tools (or gadgets) designed to gather relevant information, allowing users to identify the cause of performance issues, crashes, or other anomalies. Notably, its independence from Kubernetes enables users to employ it also for debugging control plane issues.
These tools help provide observability around many node health related problems, such as:
- Infrastructure daemon issues: NTP service down
- Hardware issues: Bad CPU, memory, or disk
- Kernel issues: Kernel deadlock, corrupted file system
- Container runtime issues: Unresponsive runtime daemon
The extension doesn't require additional outbound access to any URLs, IP addresses, or ports beyond the documented AKS egress requirements. It doesn't require any special permissions granted in Azure. It uses kubeconfig to connect to the API server to send the monitoring data collected.
Most clusters are deleted upon user request. In some cases, especially cases where you bring your own Resource Group or perform cross-RG tasks, deletion can take more time or even fail. If you have an issue with deletes, double-check that you don't have locks on the RG, that any resources outside of the RG are disassociated from the RG, and so on.
If you have issues with create and update cluster operations, make sure you don't have any assigned policies or service constraints that may block your AKS cluster from managing resources like VMs, load balancers, tags, etc.
You can, but we don't recommend it. You should perform updates when the state of the cluster is known and healthy.
If I have a cluster with one or more nodes in an Unhealthy state or shut down, can I perform an upgrade?
No, delete/remove any nodes in a failed state or otherwise from the cluster before upgrading.
Most commonly, this error arises if you have one or more Network Security Groups (NSGs) still in use that are associated with the cluster. Remove them and attempt the delete again.
Confirm your service principal hasn't expired. See AKS service principal and AKS update credentials.
Confirm your service principal hasn't expired. See AKS service principal and AKS update credentials.