I have an issue serving contents from a container (sqsh format) on worker node (remote cluster on perimeter) threw login node to my local workstation.
there are software tools requires ssh tunnel to allow access to remote services like 1. visual studio code server , 2. paraview serving GUI, 3. NVidia MATLAB container for parallel processing on GPU, 3. Jupyter Lab on nvidia docker container serving Jupyter Notebook an so on....
I can see example of ports forwarding when using docker docker run --gpus all -it --rm -p 5901:5901 -p 6080:6080 --shm-size=512M nvcr.io/partners/matlab:r2023a
how will I script this command line with srun command (what would I add to the folowing?): srun --partition mig --pty -N 1 -G 1 --cpus-per-gpu=2 --mem=60G --container-mounts $PWD:"/workspace" --container-image="./partners+matlab+r2023a.sqsh" --container-workdir="/workspace" --container-save="./ori_partners+matlab+r2023a.sqsh" bash -i
how would I perform the routing threw the slurm hosting worker node , slurm log in node to local host (tunneling to my desk top work station - ubuntu 20.04)?
Is there a "silver bullet" tunneling routing 3 levels of container - worker - login server - work station , or is it specific per each of the 4 samples I mentioned in software packages?
