I can't seem to open notebooks in Azure Machine Learning workspace

Question

I have tried changing browsers, changing accounts, recreating subscriptions.

The workspace has all the default configurations.

I tried creating a blank notebook and even that doesn't open.

I have created a compute.

the networking is public on the workspace.

Answer 1

Hi CT,

Welcome to Microsoft Q&A forum. Thank you for posting your query.

I understood that your Azure Machine Learning (AML) notebooks are not opening despite trying different browsers, accounts, and subscriptions. Since you've already ensured that a compute instance is created and your workspace is publicly accessible, let's go through a step-by-step debugging approach:

1.Check Compute Instance Status

Go to Azure Machine Learning Studio → Navigate to Compute.

Ensure the Compute Instance is in the Running state.

If it is Stopped, start the instance and try opening the notebook again.

If it is Creating, wait for it to finish before accessing notebooks.

If it is Failed, try recreating the compute.

2. Restart Compute Instance

Sometimes, restarting the compute can fix the issue.

In Azure Machine Learning Studio:

Go to Compute → Select your compute instance.

Click Stop and wait until it fully stops.

Click Start and try opening the notebook again.

3. Check Notebook Storage Location

Go to Azure Machine Learning Studio → Manage Storage.

Ensure that the storage account associated with your workspace is accessible.

If you see permission issues, ensure your account has at least Storage Blob Data Reader access to the storage.

4. Check Networking Issues

Go to Azure Machine Learning Studio → Manage Networking.

If Private Link is enabled, ensure your machine has the necessary network access.

If Public Networking is enabled, check your firewall settings.

5. Manually Open Jupyter Notebook

If the UI is not working, try launching Jupyter manually:

SSH into your compute instance:

az ml compute ssh --name <COMPUTE_NAME> --resource-group <RESOURCE_GROUP> --workspace-name <WORKSPACE_NAME>

 Start Jupyter manually:

jupyter notebook --ip=0.0.0.0 --port=8888 --allow-root

 Try accessing it via http://<compute-ip>:8888 in your browser.

6. Finally, if above does not work

Recreate Compute Instance:

Delete the compute instance.

Recreate it with a different VM size.

Try opening the notebook again.

Hope this helps. Do let us know if you any further queries.  

------------- 

If this answers your query, do click Accept Answer and Yes for was this answer helpful.

Thank you.