Saving a python notebook ipynb file takes more than 40mins via VSCode against Azure ML Compute Instance

Does this issue occur when all extensions are disabled?: Yes/No

- VS Code Version:
- OS Version:

## Steps to Reproduce:

We got a ticket internally on Azure machine learning compute instance team, as to why file save is taking so much time via VS Code.

### Details
1. Workspace => ADNOC-Digi-MLWS
2. Compute Name => ADNOC-GPU-ADITI-ARYA or ADNOC-GPU-HimaniK
3. Notebook file size => 170MB

Is the issue reproducible via any other method?
=> No. Customer confirmed that Jupyter lab works fine and saving file via vim works as well and are saved under 2s.

We evaluated and reached to below conclusions:
1. Issue is not with azure ml workspace storage or their workspace settings
2. Issue is not with Azure ML extension for VS Code either. Team confirmed that they help vs code to establish initial connection with compute instance. Once that's done they dont control how vs code is responsible for how it uses that connection.

Above inputs, tells that the issue is likely with how VS Code interacts with Compute Instance.

Could you please help us to share details on how to proceed with this ticket? Please reach out to ayushmishra@microsoft.com for any customer specific information.

Thanks