AzureML Compute page fails to load: Service temporarily unavailable

Question

AzureML Compute page fails to load: Service temporarily unavailable

Daniel Madrigal Diaz 10 Microsoft Employee

When navigating to Azure ML → Compute → Compute instances, the portal fails to retrieve the list of computes and displays the following error:

Failed to get the list of computes. ServiceError: Service temporarily unavailable. Please try again later.

This occurs directly in the Azure ML Studio UI and prevents viewing or managing existing compute instances. The issue appears to be service-side rather than configuration-related.

Karnam Venkata Rajeswari 1,150 Reputation points Microsoft External Staff Moderator

2026-04-06T18:59:17.8733333+00:00
Hello Daniel Madrigal Diaz,

This issue has been reviewed carefully in the context of how Azure Machine Learning Studio retrieves and manages compute resources through the control plane. The observed behavior where the Compute section in Azure ML Studio intermittently fails to load and returns a “Service temporarily unavailable” message has been seen in previous instances where the studio experience was temporarily unable to enumerate compute resources, despite no underlying workspace or compute configuration changes.

In such scenarios, the Studio interface depends on multiple backend services to fetch compute metadata. When any of these services experience a short‑lived disruption or partial degradation, the UI may not be able to render the list of existing compute instances or clusters. This typically presents as a transient retrieval failure rather than a persistent configuration issue, and similar behavior has historically resolved once backend availability was restored.

To continue day‑to‑day operations while the Studio interface stabilizes, Please consider following the bellow steps :

Using CLI to validate backend availability

Please run the following command to confirm whether compute resources are accessible outside of the Studio UI:

az ml compute list --resource-group <resource-group> --workspace-name <workspace-name>

A successful retrieval through CLI generally confirms that the issue is limited to the Studio experience and not the compute resources themselves.

Retrying the Studio experience after a short interval

Usually , a temporary backend unavailability often self‑recovers without manual intervention.

So ,refreshing the Azure ML Studio Compute page after some time may allow the compute list to load normally.

Checking Service Health for targeted advisories

The Azure Service Health provides subscription‑scoped updates for service incidents that may not appear on the global status page.

So , reviewing service health can help confirm whether related advisories are active for the region or subscription

Consider continuing compute operations using alternate access paths - to ensure ongoing work is not blocked during transient UI‑side disruptions. ,compute start, stop, and deletion operations can be continued through CLI or SDK even when the Studio UI is temporarily unavailable.

The following references might be helpful , please check them out

What is Azure Service Health? - Azure Service Health | Microsoft Learn

Thank you

Daniel Madrigal Diaz 10 Microsoft Employee

I’m getting a similar error when running the command below:

az ml compute list --resource-group <resource-group> --workspace-name <workspace-name>

{
  "error": {
    "code": "ServiceError",
    "severity": null,
    "message": "InternalServerError",
    "messageFormat": null,
    "messageParameters": null,
    "referenceCode": null,
    "detailsUri": null,
    "target": null,
    "details": [],
    "innerError": null,
    "debugInfo": null,
    "additionalInfo": null
  },
  "correlation": {
    "operation": "91b89439588bd979c1680c2a1c452b76",
    "request": "961eed95f896a659"
  },
  "environment": "eastus",
  "location": "eastus",
  "time": "2026-04-07T20:00:04.4962553+00:00",
  "componentName": "machinelearningcompute",
  "statusCode": 500
}