Skip to content

Microsoft.ContainerService (AKS): agentPoolProfiles showing wrong count value #12078

Open
@bcchagas

Description

@bcchagas

Problem

The REST API endpoint: /subscriptions/{{subscription}}/providers/Microsoft.ContainerService/managedClusters returns a value different from /subscriptions/{{subscription}}/resourceGroups/{{rg}}/providers/Microsoft.ContainerService/managedClusters/{{clustername}}

Description

I've created a Grafana dashboard to monitor when a nodepool reaches it's maxCount value.
Since then, I've noticed the REST API returns the wrong count number for some agentPoolProfiles whereas the Portal and az-cli returns the correct values. I have clusters stuck with the wrong count number for days, so it's not just a delay on the REST API side.

I checked it applies to the REST API versions 2020-09-01, 2020-06-01 and 2020-03-01.
It happened on clusters running Kubernetes versions 1.17.9 and 1.18.10.

Endpoint

https://management.azure.com/subscriptions/{{subscription}}/providers/Microsoft.ContainerService/managedClusters

Example

REST API Request

curl -X GET \ 'https://management.azure.com/subscriptions/{{subscription}}/providers/Microsoft.ContainerService/managedClusters?api-version=2020-09-01&skipToken=0'

REST API Response

          "agentPoolProfiles": [
                    {
                        "name": "mynodepool",
                        "count": 10,
                        "vmSize": "Standard_D8s_v3",
                        "osDiskSizeGB": 512,
                        "vnetSubnetID": "/subscriptions/{{subscription}}/resourceGroups/{{rg}}/providers/Microsoft.Network/virtualNetworks/{{vnet}}/subnets/{{subnet}}",
                        "maxPods": 110,
                        "type": "VirtualMachineScaleSets",
                        "maxCount": 10,
                        "minCount": 5,
                        "enableAutoScaling": true,
                        "provisioningState": "Succeeded",
                        "orchestratorVersion": "1.18.10",
                        "enableNodePublicIP": false,
                        "nodeLabels": {
                            "app": "myapp"
                        },
                        "mode": "User",
                        "osType": "Linux"
                    },

AZ-CLI Request

az aks show --name {{clustername}} --resource-group {{rg}}

AZ-CLI Response

"agentPoolProfiles": [
    {
      "availabilityZones": null,
      "count": 5,
      "enableAutoScaling": true,
      "enableNodePublicIp": false,
      "maxCount": 10,
      "maxPods": 110,
      "minCount": 5,
      "mode": "User",
      "name": "mynodepool",
      "nodeImageVersion": "AKSUbuntu-1804-2020.10.21",
      "nodeLabels": {
        "app": "myapp"
      },
      "nodeTaints": null,
      "orchestratorVersion": "1.18.10",
      "osDiskSizeGb": 512,
      "osType": "Linux",
      "provisioningState": "Succeeded",
      "proximityPlacementGroupId": null,
      "scaleSetEvictionPolicy": null,
      "scaleSetPriority": null,
      "spotMaxPrice": null,
      "tags": null,
      "type": "VirtualMachineScaleSets",
      "upgradeSettings": null,
      "vmSize": "Standard_D8s_v3",
      "vnetSubnetId": 
"/subscriptions/{{subscription}}/resourceGroups/{{rg}}/providers/Microsoft.Network/virtualNetworks/{{vnet}}/subnets/{{subnet}}"
    }

Metadata

Metadata

Assignees

No one assigned

    Labels

    Container ServiceService AttentionWorkflow: This issue is responsible by Azure service team.questionThe issue doesn't require a change to the product in order to be resolved. Most issues start as that

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions