Kernel is not starting in VS code

1

Dear HPC team,

I have been facing issues while starting the kernel in a VS code notebook like this:

image
image848×237 18.6 KB

This is how the output log looks like

error 17:20:20.963: Error in execution (get message for cell) [a [Error]: Unable to start Kernel 'clusterlens (Python 3.9.12)' due to connection timeout. 
View Jupyter [log](command:jupyter.viewOutput) for further details.
	at /home/users/s/sahasay2/.vscode-server/extensions/ms-toolsai.jupyter-2022.9.1303220346/out/extension.node.js:2:2196796
	at runMicrotasks (<anonymous>)
	at runNextTicks (node:internal/process/task_queues:61:5)
	at listOnTimeout (node:internal/timers:528:9)
	at processTimers (node:internal/timers:502:7)
	at async x.launch (/home/users/s/sahasay2/.vscode-server/extensions/ms-toolsai.jupyter-2022.9.1303220346/out/extension.node.js:2:2196867)] {
  category: 'timeout',
  kernelConnectionMetadata: {
    kind: 'startUsingPythonInterpreter',
    kernelSpec: {
      specFile: '/home/users/s/sahasay2/.vscode-server/extensions/ms-toolsai.jupyter-2022.9.1303220346/temp/jupyter/kernels/python3912jvsc74a57bd01883db892102973d874819fd14de5c8b760f245e761a043f10233516056de8ab/kernel.json',
      interpreterPath: '/home/users/s/sahasay2/.conda/envs/clusterlens/bin/python',
      isRegisteredByVSC: 'registeredByNewVersionOfExt',
      name: 'python3912jvsc74a57bd01883db892102973d874819fd14de5c8b760f245e761a043f10233516056de8ab',
      argv: [Array],
      language: 'python',
      executable: 'python',
      display_name: "Python 3.9.12 ('clusterlens': conda)",
      metadata: [Object],
      env: {},
      interrupt_mode: undefined
    },
    interpreter: {
      id: '/home/users/s/sahasay2/.conda/envs/clusterlens/bin/python',
      sysPrefix: '/home/users/s/sahasay2/.conda/envs/clusterlens',
      envType: 'Conda',
      envName: 'clusterlens',
      envPath: [p],
      architecture: 3,
      sysVersion: '3.9.12 (main, Apr  5 2022, 06:56:58) \n[GCC 7.5.0]',
      version: [Object],
      companyDisplayName: 'Anaconda, Inc.',
      displayName: "Python 3.9.12 ('clusterlens': conda)",
      detailedDisplayName: "Python 3.9.12 ('clusterlens': conda)",
      uri: [p]
    },
    id: '.jvsc74a57bd01883db892102973d874819fd14de5c8b760f245e761a043f10233516056de8ab./home/users/s/sahasay2/.conda/envs/clusterlens/python./home/users/s/sahasay2/.conda/envs/clusterlens/python.-m#ipykernel_launcher'
  }
}

Could you please help to understand the problem? FYI, the clusterlens environment completely works from terminal.

Thank & Regards,
Sayan

2

Hi,

Could you give me more details about how you launch your job (your sbatch)? :slight_smile:

3

Hi, here I am just trying to use VS code using the remote-ssh extension. When I am running a jupyter-notebook. The kernel is not starting. :confused:

4

Hi, we remind you that you have to launch your Jupypter session on a compute node and not directly on the login node.

Here is a tutorial I presented during the latest hpc-lunch.

For our understanding: what is the point to launch Jupyter through vscode on the login node or compute node?