Unable to deploy any models in Microsoft Foundry (Error 715-123420)

MadCuts AI 20 Reputation points
2026-05-28T18:39:35.6966667+00:00

I am unable to deploy any model in Microsoft Foundry. Every attempt results in the same error banner:

715-123420: An error occurred. Please reach out to support for additional assistance.

I would greatly appreciate urgent assistance in resolving this issue, as it is preventing all model deployments. Please advise if any additional information, logs, or permissions details are needed to troubleshoot this problem effectively.

Microsoft Foundry
Microsoft Foundry

A unified Azure platform for creating and managing AI models, agents, and applications with built‑in enterprise security, monitoring, and governance

0 comments

2 answers

Sort by: Oldest
Most helpful Newest Oldest

  1. AI answer

    2026-05-28T18:39:51.69+00:00

    Error 715-123420 is a generic deployment failure banner. Use the following steps to isolate the root cause and collect the details support will need.

    1. Check deployment logs in Foundry
      • In Microsoft Foundry, open the affected project.
      • Go to Models + endpoints (or the deployment list) and open the failing deployment.
      • Review the deployment logs for a more specific error (for example, quota exceeded, authorization failed, model not available, or environment/image build issues).
    2. Verify common deployment blockers
      • Quota / capacity
        • In the Foundry portal, open Management center → Quota for the Foundry/Foundry Models resource and confirm that:
          • The selected model and SKU have available quota in the target region.
          • Provisioned capacity (PTUs) is available if using provisioned deployments.
        • If logs show quota or capacity errors, either request a quota increase or try a different region/SKU as described in the quota guidance.
      • Authorization / RBAC
        • If logs show authorization failures, ensure the identity performing the deployment has the required role on the Foundry or Cognitive Services resource (for example, Cognitive Services Contributor or the documented role for that deployment path).
      • Model / region availability
        • If logs indicate the model is not available, confirm that the model supports the selected deployment type and region. Choose a supported region or deployment type if needed.
    3. If using classic (hub-based) Foundry projects
      • For hub-based (classic) projects where deployments are backed by Azure Machine Learning online endpoints:
        • Confirm the hub’s managed identity has the Azure ML Data Scientist role on the resource group:
          1. In Foundry (classic), go to the project → Management center → Project → Overview.
          2. Open the linked resource group in the Azure portal.
          3. Go to Access control (IAM) → Add role assignment.
          4. Assign Azure ML Data Scientist to the Machine learning online endpoint managed identity for the endpoint.
          5. Return to Foundry and retry the deployment.
      • If the error text mentions ImageBuildFailure or timeouts waiting for an environment image:
        • Retrieve the environment build log either:
          • From the storage account and container path indicated in the error, via Data storage → Containers and navigating to the image_build_aggregate_log.txt, or
          • From Azure Machine Learning studio: Endpoints → [endpoint] → Environment and check the build log.
        • Use the detailed error in that log to fix missing packages, base image issues, or other environment problems, then redeploy.
    4. Check for generic provisioning failures
      • If deployments stay in provisioning or time out:
        • Confirm the hub’s managed network settings allow access to required resources.
        • Verify the model ID is valid and still available.
        • Try a different VM SKU or reduce instance count if using managed compute.
    5. Collect information for support If the banner persists for all models and no specific cause is visible in logs, collect the following before opening a support ticket (as the banner suggests):
      • Subscription ID and Foundry/Foundry Models resource name and region.
      • Project name and deployment name(s).
      • Exact time (UTC) of a recent failed deployment.
      • Screenshots or copies of:
        • The 715-123420 banner.
        • Any detailed error messages from the deployment logs.
        • Quota view for the target model/SKU/region.
      • For classic projects: any environment build log excerpts showing ImageBuildFailure or related errors.

    Then open an Azure support request so the backend team can correlate the 715-123420 error with internal telemetry and clear any stuck quota/capacity or platform issues.

    References:

    AI-generated content may be incorrect. Read our transparency notes for more information.

    Was this answer helpful?

  2. Jerald Felix 18,680 Reputation points Volunteer Moderator
    2026-05-29T04:42:35.4266667+00:00

    Hello MadCuts AI,

    Greetings! Thanks for raising this question in Q&A forum.

    The error 715-123420 in Microsoft Foundry is a generic deployment failure that can be triggered by a few different root causes most commonly insufficient quota, missing role permissions, or a regional availability issue with the model you're trying to deploy. Let's work through this together.

    Step 1: Check the Deployment Logs for a Specific Error

    In Microsoft Foundry, open your project and go to Models + endpoints. Click on the failing deployment and review the logs. This will usually reveal a more specific error like quota exceeded, authorization failed, or model not available — which will guide the next steps.

    Step 2: Check Your Quota

    Go to Management center → Quota in the Foundry portal. Confirm that the model and SKU you are trying to deploy have available quota in your target region. If you are using provisioned deployments (PTUs), make sure provisioned capacity is also available. If quota is the issue, you can request an increase or try deploying in a different region.

    Step 3: Verify Your Permissions (RBAC)

    If the logs show an authorization error, check that the account performing the deployment has the right role. At minimum, you'll need the Cognitive Services Contributor role on the Foundry or Azure AI resource. Go to the Azure Portal, navigate to your resource, click Access control (IAM), and verify your role assignment.

    Step 4: Confirm Model and Region Availability

    Not all models are available in all regions. Check that the model you are trying to deploy supports the deployment type and region you have selected. If not, switch to a supported region and retry.

    Step 5: For Classic (Hub-based) Projects — Check Managed Identity Role

    If you are using a classic hub-based Foundry project, the hub's managed identity needs the Azure ML Data Scientist role on the resource group. Go to your project in Foundry → Management center → Project → Overview, open the linked resource group in the Azure Portal, go to Access control (IAM), and assign that role to the Machine learning online endpoint managed identity. Then retry the deployment.

    Step 6: If Nothing Works — Open a Support Ticket

    Since the error banner itself recommends contacting support, please collect the following before raising a ticket your Subscription ID, Foundry resource name and region, project and deployment name, exact UTC time of the failure, screenshots of the error banner, and any logs showing specific errors. Then open a support request at https://aka.ms/azuresupport so the backend team can correlate the error with internal telemetry on their end.

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

    Best Regards,

    Jerald Felix.

    Was this answer helpful?

    0 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.