I am accessing jupyterhub notebook through browser as it is running on a remote server(using nginx so that jupyterhub is accessible to client machines).When I access the notebook a kernel error is shown at the top of the screen. Jupyterhub and jupyter notebook is running on a kubernetes cluster and the jupyter notebook is a custom built docker image suitable for python programming. I am very new into python/jupyter hub/notebook, please suggest how to resolve the kernel error.
1
votes
If you click on the "Kernel error" it should show a log. Could you please post the error log?
– alec_djinn
on clicking on the kernel error it says, failed to start the kernel. not found. And when i do ´jupyter kernel spec list´ , I get the following response:Available kernels: python3 /usr/local/share/jupyter/kernels/python3. Following are the content of my kernel.json file { "display_name": "python3", "language": "python", "argv": [ "/usr/bin/python3", "-m", "ipykernel_launcher", "-f", "{connection_file}" ] }. python3 is present in the path mentioned in the kernel.json file. I am running jupyterhub and notebook on a kubernetes cluster which is setup on ubuntu 16.04.
– Nitesh Ratnaparkhe
@alec_djinn can you plz help?
– Nitesh Ratnaparkhe
Unfortunately I can't without a proper error log. If that is the only info you have I don't know where to start. You can set a bounty (probably you don't have this right yet) to attract more attention to this post. Hopefully someone more skilled than me can figure it out.
– alec_djinn
2 Answers
1
votes
Please check kernel.json
file, check if the path to the Python
executable is correct
for example my kernel.json
looks like this, in my case the Python binary is located at
C:/Users/Vinod/Anaconda/envs/env_name/python
Check the path in your case
{
"argv": [
"C:/Users/Vinod/Anaconda/envs/env_name/python",
"-m",
"ipykernel_launcher",
"-f",
"{connection_file}"
],
"display_name": "Python 3",
"language": "python"
}
Note: env_name refers to the anaconda environment you are using