SantaHub/JupyterhubCUDA

JupyterhubCUDA

JupyterhubCUDA

Cloning the Datalab Image

docker run -d --gpus all -p 8000:8000 -p 8787:8787 --name datalab_image austinputhen/datalab:version2 

docker exec -it IMAGE_ID bash

GPU Activation

• install the proprietary and tested NVIDIA Driver in additional drivers option in software & update

• Reboot to take effect

• nvidia-smi should list the GPUs you have.

• since we are using docker images we dont have to install CUDA on host machine

• install for docker

sudo apt update
sudo apt install -y git curl docker
distribution=$(. /etc/os-release;echo $ID$VERSION_ID) \
  && curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add - \
  && curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudo tee /etc/apt/sources.list.d/nvidia-docker.list && sudo apt update
sudo apt install -y nvidia-docker2
sudo systemctl restart docker

Test with

sudo docker run --rm --gpus all nvidia/cuda:11.0-base nvidia-smi

Get tensorflow image

Ref : https://hub.docker.com/r/tensorflow/tensorflow/

docker run --rm -it --gpus all -p 8000:8000 -p 8787:8787  --name datalab  tensorflow/tensorflow bash

Installing Jupyterhub

 apt install npm
 pip install jupyterhub
 npm install -g configurable-http-proxy

Verifying CUDA installation in Jupyterhub

import tensorflow as tf
tf.config.list.physical_devices('GPU')
tf.test.is_build_with_cuda()

Tensorflow

import tensorflow as tf
tf.config.list_physical_devices('GPU')
tf.test.is_gpu_available()

Torch

import torch
torch.cuda.is_available()
torch.cuda.device_count()
torch.cuda.get_device_name(0)

Install R Server

sudo apt install -y r-base gdebi-core wget
wget https://download2.rstudio.org/server/bionic/amd64/rstudio-server-2021.09.1-372-amd64.deb
sudo gdebi rstudio-server-2021.09.1-372-amd64.deb

sudo rstudio-server verify-installation

Login to RServer : http://localhost:8787/

Adding new User

adduser student

Start Services

Add this to bashrc.

rstudio-server restart
jupyterhub

Docker Commands

Run Docker image

docker run -d --gpus all -p 8000:8000 -p 8787:8787 --name dl1 datalab/jupyterhub_cuda_rserver jupyterhub

Execute a command on the container

docker exec -it cd0c0aadcd96 bash

Copy stuffs

docker cp <containerId>:/file/path/within/container /host/path/target
docker cp ~/Documents/JupyterhubCUDA/bash_startup 6586b8b05fd8:/root/.bash_startup

Saving docker images

docker commit <IAMGE_ID/IMAGE_NAME> <NEW-IMAGE-NAME>
docker commit 0fd419771dcd jupyterHubDL