I'm trying to start a Spyder kernel from a docker container and connect the Spyder IDE on the host to this kernel on the container (I tried follow this post, and the instructions on Spyder's site - Connect Spyder to a console in a docker container on a remote host). However, I encountered the following problems:
- Getting access to the kernel-pid.json file which is in the container. I tried to map the ~/.local/shared/jupyter/runtime folder on the host to the one on the container but I'm not sure it's the right way to do it (maybe I need to connect it via SSH, like in remote connection).
- If I do map this folder on the host to the container, I need to give permissions (using sudo chmod) to the kernal-pid.json file, or else it says it doesn't have the appropriate permissions when I tried to connect to the kernel on the docker container.
- If I give the appropriate permissions, it looks like it goes well, but then the process is stuck forever in "connecting to kernel" message (all of this is inside the Spyder IDE).
I have a docker version of 19.03, which have anaconda3 and python 3.7 installed on it, and on the host I also have anaconda3 (to run the Spyder IDE).