search

microsoft / vscode-tools-for-ai

Azure Machine Learning for Visual Studio Code, previously called Visual Studio Code Tools for AI, is an extension to easily build, train, and deploy machine learning models to the cloud or the edge with Azure Machine Learning service.
Other
324 stars 94 forks source link

Can't connect to remote machine - keep getting this error #2331

Closed felicityallen closed 6 months ago

felicityallen commented 6 months ago

Does this occur consistently? Yes at the moment, but No as in it was doing it a week or two ago, then stopped for a few days, and is now doing it again. Not sure what's changed.

Repro steps:

  1. From Machine Learning studio in Azure, click on 'VS Code Desktop' for compute node.

Action: Resolver.resolve Error type: ie Error Message: Failed to connect to target. Jupyter is required to connect to the remote target. Please ensure Jupyter is installed and running on the remote machine and retry connecting from the extension or the ML Studio. (Error: Invalid response: 502 Bad Gateway)

Version: 0.46.0 OS: darwin OS Release: 23.2.0 Product: Visual Studio Code Product Version: 1.87.0 Language: en

Call Stack ``` ie.NotAvailable extensionHostProcess.js:148:8631 w.handle502BadGateway extension.js:2:1611528 w.handleError extension.js:2:1608225 o.value extension.js:2:1532109 extension.js:2:1624225 ```
shsuman commented 6 months ago

@felicityallen Ae you able to access Jupyter Lab from Azure ML Studio when this error happens ?

felicityallen commented 6 months ago

No, it seems to time out.On 13 Mar 2024, at 06:42, Shantnu Suman @.***> wrote: @felicityallen Ae you able to access Jupyter Lab from Azure ML Studio when this error happens ?

—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you were mentioned.Message ID: @.***>

shsuman commented 6 months ago

@felicityallen In that case, there seems to be a problem with the Jupyter running on Compute instances. VS Code depends on Jupyter to be up and running before we can make a connection. Please open an issue with Azure ML Studio directly

felicityallen commented 6 months ago

Why does VScode need Jupyter to be running when I'm not using Jupyter? I will try logging a request with ML studio.

shsuman commented 6 months ago

@felicityallen We need Jupyter terminal to install and run VS Code server on the Compute instances