When will new SKUs be available in NAP?

[Chrisbattarbee]

What karpenter features are relevant?

Hey there, we're looking to use Standard_F8amds_v7 with NAP and I can see that the SKU is known here:

karpenter-provider-azure/pkg/providers/instancetype/known_skus.yaml

Line 3194 in 9031fb9

- name: Standard_F8amds_v7

Any idea when this will actually be available with NAP in Azure, we have a similar situation with the laos v4 SKUs too

karpenter-provider-azure/pkg/providers/instancetype/known_skus.yaml

Line 3593 in 9031fb9

- name: Standard_L4aos_v4

How should the docs be improved?

It would be great to know what is available in azure NAP and what is not, what the process / timeline is for things being added here and being made available in azure

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

[matthchr]

Both of the above SKUs were added #1606. The above change is rolling out for NAP now and should be available in all public production regions as of yesterday.

Can you try again and see if they work for you now, and if not share your region (required) and cluster (ideally) details with us?

[benritchie]

Matthew - I've also hit this issue.
Just tried to repo, and still seeing it. Details:

• region - west europe.
• VM type: Standard_D2ads_v7
• no availability zones selected. (I know these VM's are only available in 2 / 3 of the west europe availability zones).
• I've confirmed with a v6 VM with all other settings identical, which works.
  Happy to share cluster details privately if useful.

[matthchr]

According to my metrics, there are still a few clusters in WestEurope that don't have 1.10.2 version, either just due to it taking a bit for Karpenter to get updated in that region or (more likely) maintenanceWindows defined on those clusters that prevent updating Karpenter for a bit longer than average.

I am matthchr on k8s slack, if you want to share your cluster details with me (just subscriptionID + name is enough) I can see if I can understand why.

[Chrisbattarbee]

Hey @matthchr , looks like we also dont have it in UKSouth. I have a cluster in central america, I can try it on next. The result of describing the nodepool is below. Will also hop on k8s slack and dm you cluster details.

Name:         clickhouse
Namespace:    
Labels:       app.kubernetes.io/component=node-pool
              app.kubernetes.io/managed-by=Helm
              app.kubernetes.io/name=nap-nodepools
Annotations:  karpenter.sh/nodepool-hash: 7857275499387201084
              karpenter.sh/nodepool-hash-version: v3
              meta.helm.sh/release-name: metoro-nap-nodepools
              meta.helm.sh/release-namespace: metoro
API Version:  karpenter.sh/v1
Kind:         NodePool
Metadata:
  Creation Timestamp:  2026-04-26T17:36:30Z
  Generation:          19
  Resource Version:    5260447
  UID:                 da8d4106-941e-4dfd-8ce9-5cd1f49bfc6d
Spec:
  Disruption:
    Budgets:
      Nodes:               10%
    Consolidate After:     5m
    Consolidation Policy:  WhenEmpty
  Limits:
    Cpu:     128
    Memory:  1Ti
  Template:
    Metadata:
      Labels:
        metoro.io/workload:  clickhouse
        Workload:            clickhouse
    Spec:
      Expire After:  Never
      Node Class Ref:
        Group:  karpenter.azure.com
        Kind:   AKSNodeClass
        Name:   clickhouse
      Requirements:
        Key:       kubernetes.io/os
        Operator:  In
        Values:
          linux
        Key:       karpenter.sh/capacity-type
        Operator:  In
        Values:
          on-demand
        Key:       karpenter.azure.com/sku-name
        Operator:  In
        Values:
          Standard_F8amds_v7
        Key:       topology.kubernetes.io/zone
        Operator:  In
        Values:
          uksouth-1
          uksouth-2
          uksouth-3
      Startup Taints:
        Effect:  NoSchedule
        Key:     metoro.io/clickhouse-nvme-bootstrap
        Value:   pending
      Taints:
        Effect:  NoSchedule
        Key:     workload
        Value:   clickhouse
Status:
  Conditions:
    Last Transition Time:          2026-04-26T17:36:30Z
    Message:                       
    Observed Generation:           19
    Reason:                        NodeClassReady
    Status:                        True
    Type:                          NodeClassReady
    Last Transition Time:          2026-04-26T17:36:30Z
    Message:                       
    Observed Generation:           19
    Reason:                        ValidationSucceeded
    Status:                        True
    Type:                          ValidationSucceeded
    Last Transition Time:          2026-04-27T10:05:14Z
    Message:                       object is awaiting reconciliation
    Observed Generation:           19
    Reason:                        AwaitingReconciliation
    Status:                        Unknown
    Type:                          NodeRegistrationHealthy
    Last Transition Time:          2026-04-28T19:35:36Z
    Message:                       
    Observed Generation:           19
    Reason:                        Ready
    Status:                        True
    Type:                          Ready
  Node Class Observed Generation:  5
  Nodes:                           2
  Resources:
    Cpu:                  32
    Ephemeral - Storage:  257834976Ki
    hugepages-1Gi:        0
    hugepages-2Mi:        0
    Memory:               131684312Ki
    Nodes:                2
    Pods:                 60
Events:
  Type     Reason                     Age                 From       Message
  ----     ------                     ----                ----       -------
  Normal   Ready                      2m28s               karpenter  Status condition transitioned, Type: Ready, Status: True -> Unknown, Reason: ReconcilingDependents, Message: NodeClassReady=True
  Normal   Ready                      47s (x7 over 34h)   karpenter  Status condition transitioned, Type: Ready, Status: Unknown -> True, Reason: Ready
  Normal   Ready                      47s (x2 over 34h)   karpenter  Status condition transitioned, Type: Ready, Status: True -> Unknown, Reason: ReconcilingDependents, Message: ValidationSucceeded=True, NodeClassReady=True
  Warning  NoCompatibleInstanceTypes  21s (x19 over 34h)  karpenter  NodePool requirements filtered out all compatible available instance types

[matthchr]

I just realized (thanks @Chrisbattarbee) that the issue is the PR I linked above is not in 1.10.2. 1.10.2 is everywhere in prod now but doesn't actually contain the fix.

I'll talk with the team about when we're doing a new release - I think the plan is to cut one this week.