Unable to create Logic App from Microsoft AI Foundry due to Compute quota showing as zero

Tariq D 0 Reputation points
2026-06-29T16:34:48.2066667+00:00

Issue Summary

I am attempting to create a connection from Microsoft AI Foundry to an Azure service. During the connection setup, Microsoft AI Foundry automatically provisions a Logic App in the background. However, the deployment consistently fails.

Error

The deployment fails with the following error:

Failed to create Logic App deployment. Operation cannot be completed without additional quota.

Current Limit (Total VMs): 0 Current Usage: 0 Amount required for this deployment: 1 (Minimum)

Troubleshooting Performed

  • Verified that the required Azure Resource Providers are registered, including:
  • Microsoft. Compute * Microsoft. Logic * Microsoft. Web

Attempted to increase the Compute quota for the target region. The portal indicated that the quota request was successful*. * Waited 10–20 minutes and refreshed the portal multiple times. The quota still shows as 0*, and the updated limit is not reflected. * Attempted the deployment again from Microsoft AI Foundry, but the same error persists. * I am also unable to provision Virtual Machines, suggesting this may be a subscription-level Compute quota issue rather than an AI Foundry issue. In Subscriptions → Usage + Quotas*, no Compute quota information is displayed.

Could you please help me investigate why the Compute quota remains at zero despite the successful quota increase request and advise how to resolve this issue? Thank you.

Azure OpenAI in Foundry Models
0 comments No comments

1 answer

Sort by: Most helpful
  1. Sangeetha Kesavan 20 Reputation points
    2026-06-29T16:51:51.66+00:00

    Hello Tariq D,

    Thanks for raising this question in the Q&A forum.

    The root cause here is a regional compute quota of zero on your subscription, which prevents both the Logic App provisioned by AI Foundry and standalone Virtual Machine deployments from succeeding. When Microsoft AI Foundry creates a Logic App connection in the background, it provisions a Logic App Standard resource that runs on an underlying App Service / Workflow Standard plan. This plan requires at least 1 vCPU worth of compute quota in the target region. If the quota for that region sits at zero, the deployment fails with exactly the error you are seeing even if the quota increase request was submitted successfully.

    There are a few common reasons why the quota still shows as zero after a successful request, and the steps below address each one.

    1. Confirm the correct quota provider is being checked — The "Usage + Quotas" blade under Subscriptions sometimes displays only classic compute quotas (Microsoft.Compute classic / Cloud Services classic), which are separate from the quotas that govern App Service and Logic Apps Standard. Navigate to the Azure portal, search for "Quotas" in the top search bar, and open the dedicated Quotas blade. Set the Provider filter to Microsoft.Web and filter by your target region. Look for entries such as "Standard VMs" or "Workflow Standard" for your region. This is the quota that the Logic App deployment is drawing against, and it may show zero even if your Microsoft.Compute classic quota appears fine.
    2. Verify the quota increase targeted the correct region — Azure quotas are enforced per subscription per region. If your quota increase request was submitted for a different region than the one AI Foundry is deploying into, the target region will still show zero. Confirm which region the Logic App is being deployed to (visible in the deployment error details or the AI Foundry connection settings), then recheck the Quotas blade filtered to that specific region.
    3. Request the quota increase through the correct path — If the Microsoft.Web quota for your region is zero, raise a new support request specifically for it. In the Azure portal, go to Help + support, select New support request, set Issue type to Service and subscription limits (quotas), and set Quota type to App Service. In the request details, specify the region, the SKU (Workflow Standard or Standard), and request a new limit of at least 1. Submit the request and wait for the confirmation email indicating it has been applied, not just submitted.
    4. Check for a spending limit on the subscription — If this is a free trial, Visual Studio, or MSDN subscription, Azure applies a spending limit that can prevent quota increases from taking effect even after approval. Go to Subscriptions, select your subscription, and check for an active spending limit under the Overview blade. If one is present, you will need to remove it before quota increases can be reflected. Note that free trial subscriptions are generally not eligible for quota increases through the self-service path and require a support ticket.
    5. Try a different region as an immediate workaround — If you need to proceed while waiting for the quota increase to be processed, check whether the AI Foundry connection setup allows you to select a different deployment region. Regions such as East US 2, West US 2, or West Europe frequently have available capacity for new subscriptions and may allow the Logic App to provision successfully without a quota increase.
    6. Escalate to Azure Support if the quota increase is not reflected after 30 minutes — Approved quota increases typically propagate within a few minutes. If the Quotas blade still shows zero after 30 minutes following the approval confirmation, open a support request under Issue type: Technical, Service: Subscription Management, and describe that an approved quota increase is not being reflected. Include your subscription ID, the region, the provider (Microsoft.Web), and the date and time of the approved request.

    If this answer helps you kindly accept the answer which will help others who have similar questions.

    Regards,

    Sangeetha K

    Was this answer helpful?

    0 comments No comments

Your answer

Answers can be marked as 'Accepted' by the question author and 'Recommended' by moderators, which helps users know the answer solved the author's problem.