VM Scale set is in a Failed ProvisioningState, because of an extension #82
Description
We've discovered there is a VM Scale set, that seems to be in "failed" ProvisioningState. The nodes are running fine, but one extension on the VMSS has been reported as failed. Unfortunately this is a managed service and we don't have much of an access to troubleshoot this further. However we checked the state of the VMSS via the API with the command below:
az vmss get-instance-view --resource-group groupName --name vmssName
"name": "AKSLinuxExtension",
"statusesSummary": [
{
"code": "ProvisioningState/succeeded",
"count": 1
},
{
"code": "ProvisioningState/failed",
"count": 1
}
]
}
],
"orchestrationServices": null,
"statuses": [
{
"code": "ProvisioningState/failed/VMExtensionHandlerNonTransientError",
"displayStatus": "Provisioning failed",
"level": "Error",
"message": "The handler for VM extension type 'Microsoft.AKS.Compute.AKS.Linux.AKSNode' has reported terminal failure for VM extension 'AKSLinuxExtension' with error message: 'Failed to download artifacts: [ExtensionDownloadError] Timeout downloading extension package. Elapsed: 0:08:04.405546 URIs tried: 2/3. Last error: [HttpError] Download failed both on the primary and fallback channels. IOError [Errno -3] Temporary failure in name resolution -- 6 attempts made]'.\r\n \r\n'Manifest download failed for the extension.
When we ran on all VMSS nodes the same command it showed that everything looks fine:
"name": "AKSLinuxExtension",
"statuses": [
{
"code": "ProvisioningState/succeeded",
"displayStatus": "Provisioning succeeded",
"level": "Info",
"message": "Successfully enabled Compute.AKS.Linux.AKSNode extension",
"time": null
}
],
"substatuses": null,
"type": "Microsoft.AKS.Compute.AKS.Linux.AKSNode",
"typeHandlerVersion": "1.56"
}
The question here is - how do we fix the error of the VMSS? Will restarting it fix the issue without a downtime, or there is another more pragmatic way to clear this error?