terraform - Create a gpu nodepool on azure with a student account - Stack Overflow

admin管理员组

文章数量:1127072

I want to follow this tutorial, explaining how one can run ollama on Azure. As a node pool this tutorial uses Standard_NC6s_v3. According to cloudprice there are some cheaper alternatives. But at NC6s_v3 and other variations I get the error NotAvailableForSubscription. Since I just have a student account. Is there any other vm size which has a GPU enabled that I could use with my subscription? It should have a decent GPU and about 90GB of RAM. I know it's a lot but I just want to run it for a couple of minutes until the inference is finished. For testing purposes a smaller variation with less GB RAM would also be good. There I can create a setup script that I could run in terraform at each start.

When i want to run the following script:

resource "azurerm_virtual_machine" "this" {
    name                  = "gpu-vm"
    location              = azurerm_resource_group.this.location
    resource_group_name   = azurerm_resource_group.this.name
    network_interface_ids = [azurerm_network_interface.this.id]
    vm_size               = "Standard_NC6s_v3"

    storage_os_disk {
        name              = "osdisk"
        caching           = "ReadWrite"
        create_option     = "FromImage"
        managed_disk_type = "Standard_LRS"
    }

    storage_image_reference {
        publisher = "Canonical"
        offer     = "UbuntuServer"
        sku       = "18.04-LTS"
        version   = "latest"
    }

    os_profile {
        computer_name  = "gpu-vm"
        admin_username = "adminuser"
        admin_password = "Password1234!"
    }

    os_profile_linux_config {
        disable_password_authentication = false
    }

    tags = {
        environment = "production"
    }
}

I get:

╷
│ Error: creating/updating Virtual Machine (Subscription: "..."
│ Resource Group Name: "gpu-rg"
│ Virtual Machine Name: "gpu-vm"): performing CreateOrUpdate: unexpected status 409 (409 Conflict) with error: OperationNotAllowed: Operation could not be completed as it results in exceeding approved standardNCSv3Family Cores quota. Additional details - Deployment Model: Resource Manager, Location: eastus, Current Limit: 0, Current Usage: 0, Additional Required: 6, (Minimum) New Limit Required: 6. Setup Alerts when Quota reaches threshold. Learn more at  . Submit a request for Quota increase at .ReactView/Parameters/%7B%22subscriptionId%..mand%22:%22openQuotaApprovalBlade%22,%22quotas%22:[%7B%22location%22:%22eastus%22,%22providerId%22:%22Microsoft.Compute%22,%22resourceName%22:%22standardNCSv3Family%22,%22quotaRequest%22:%7B%22properties%22:%7B%22limit%22:6,%22unit%22:%22Count%22,%22name%22:%7B%22value%22:%22standardNCSv3Family%22%7D%7D%7D%7D]%7D by specifying parameters listed in the ‘Details’ section for deployment to succeed. Please read more about quota limits at 
│ 
│   with azurerm_virtual_machine.this,
│   on nodepool.tf line 19, in resource "azurerm_virtual_machine" "this":
│   19: resource "azurerm_virtual_machine" "this" {

I want to follow this tutorial, explaining how one can run ollama on Azure. As a node pool this tutorial uses Standard_NC6s_v3. According to cloudprice.net there are some cheaper alternatives. But at NC6s_v3 and other variations I get the error NotAvailableForSubscription. Since I just have a student account. Is there any other vm size which has a GPU enabled that I could use with my subscription? It should have a decent GPU and about 90GB of RAM. I know it's a lot but I just want to run it for a couple of minutes until the inference is finished. For testing purposes a smaller variation with less GB RAM would also be good. There I can create a setup script that I could run in terraform at each start.

When i want to run the following script:

resource "azurerm_virtual_machine" "this" {
    name                  = "gpu-vm"
    location              = azurerm_resource_group.this.location
    resource_group_name   = azurerm_resource_group.this.name
    network_interface_ids = [azurerm_network_interface.this.id]
    vm_size               = "Standard_NC6s_v3"

    storage_os_disk {
        name              = "osdisk"
        caching           = "ReadWrite"
        create_option     = "FromImage"
        managed_disk_type = "Standard_LRS"
    }

    storage_image_reference {
        publisher = "Canonical"
        offer     = "UbuntuServer"
        sku       = "18.04-LTS"
        version   = "latest"
    }

    os_profile {
        computer_name  = "gpu-vm"
        admin_username = "adminuser"
        admin_password = "Password1234!"
    }

    os_profile_linux_config {
        disable_password_authentication = false
    }

    tags = {
        environment = "production"
    }
}

I get:

╷
│ Error: creating/updating Virtual Machine (Subscription: "..."
│ Resource Group Name: "gpu-rg"
│ Virtual Machine Name: "gpu-vm"): performing CreateOrUpdate: unexpected status 409 (409 Conflict) with error: OperationNotAllowed: Operation could not be completed as it results in exceeding approved standardNCSv3Family Cores quota. Additional details - Deployment Model: Resource Manager, Location: eastus, Current Limit: 0, Current Usage: 0, Additional Required: 6, (Minimum) New Limit Required: 6. Setup Alerts when Quota reaches threshold. Learn more at https://aka.ms/quotamonitoringalerting . Submit a request for Quota increase at https://aka.ms/ProdportalCRP/#blade/Microsoft_Azure_Capacity/UsageAndQuota.ReactView/Parameters/%7B%22subscriptionId%...command%22:%22openQuotaApprovalBlade%22,%22quotas%22:[%7B%22location%22:%22eastus%22,%22providerId%22:%22Microsoft.Compute%22,%22resourceName%22:%22standardNCSv3Family%22,%22quotaRequest%22:%7B%22properties%22:%7B%22limit%22:6,%22unit%22:%22Count%22,%22name%22:%7B%22value%22:%22standardNCSv3Family%22%7D%7D%7D%7D]%7D by specifying parameters listed in the ‘Details’ section for deployment to succeed. Please read more about quota limits at https://docs.microsoft.com/en-us/azure/azure-supportability/per-vm-quota-requests
│ 
│   with azurerm_virtual_machine.this,
│   on nodepool.tf line 19, in resource "azurerm_virtual_machine" "this":
│   19: resource "azurerm_virtual_machine" "this" {

• azure
• terraform
• ollama

Share Improve this question Follow edited Jan 9 at 8:13 Cowboy_Patrick asked Jan 8 at 18:50 Cowboy_PatrickCowboy_Patrick 744 bronze badges 2

• Have you tried any terraform script related to your requirement. @Cowboy_Patrick – Jahnavi Commented Jan 9 at 3:36
• 1 I did, I added the script and the response to the question – Cowboy_Patrick Commented Jan 9 at 8:13

Add a comment  | 

1 Answer 1

Sorted by: Reset to default 0

There will be a quota limitation when it comes to the Azure free or student related subscriptions and High-end GPUs like NC6s_v3 are a bit expensive resource.

Azure development might restrict access to these accounts and provide availability for others based on their business requirements.

Is there any other vm size which has a GPU enabled that I could use with my subscription?

Yes, there are detailed view of VM sizes with each of their family types accordingly listed in MSDoc here. You can check it through storage optimized, memory optimized, compute optimized and also GPU accelerated and so on.

N series VM's are the GPU node pool optmised VM's. Refer Microsoft Q&A for more similar information.

If still the issue persists with higher GPU node pools, then you can raise a New Quota Request under the specific subscription to work with the higher capability CPU or GPU's.

I tried the same code as you in my environment without any changes and it showed me the below terraform plan to proceed further without any Not available errors as I have a pay as you go business account.

本文标签： terraformCreate a gpu nodepool on azure with a student accountStack Overflow