You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
My colleagues are standing up a jupyterhub to allow users to select conda environments I maintain. They setup a environment to run jupyterhub out of a python 3 conda environment, and then users select a python 2.7 conda environment. The issue is that once the kernel has started, the environment has issues. In particular the PATH still points to the python 3 environment jupyterhub runs from.
The conda environements I maintain set environment variables using the
etc/conda/activate.d/env_vars.sh
mechanism. It does not appear that the jupyterhub ipython mechanism has done anything like a source activate for the environment for the kernel. How does this work?
The text was updated successfully, but these errors were encountered:
It does not appear that the jupyterhub ipython mechanism has done anything like a source activate for the environment for the kernel. How does this work?
That's correct, nb_conda_kernels is creathing a kernespec on the fly pointing to the environment where jupyter (ipykernel) is detected.
I think you are experiencing something like this: #19
My colleagues are standing up a jupyterhub to allow users to select conda environments I maintain. They setup a environment to run jupyterhub out of a python 3 conda environment, and then users select a python 2.7 conda environment. The issue is that once the kernel has started, the environment has issues. In particular the PATH still points to the python 3 environment jupyterhub runs from.
The conda environements I maintain set environment variables using the
mechanism. It does not appear that the jupyterhub ipython mechanism has done anything like a source activate for the environment for the kernel. How does this work?
The text was updated successfully, but these errors were encountered: