Scipion-docker

This project has been moved to GitHub, and this repository now serves only as an archive of the old version!

Deploy Scipion using IM

1. Go to IM-web and log in with your account

2. Make sure you have access to the desired platform

You can open a list your Crendentials on the left menu. If necessary, add a new one by clicking the "Add +" button and entering correct information.

3. Paste the RADL/TOSCA recipe into the new IM topology

Click the Topologies button in the left menu and the Add + button at the top.

Paste the RADL/TOSCA recipe from the im-topology.yml file and save the topology.

Please change default password for VNC at the beginning of the Ansible section!

You can edit the recipe if needed. More info about the IM: https://imdocs.readthedocs.io/en/latest/intro.html

4. Launch your topology

In the list of topologies launch the one you just created. You will be redirected to the Infrastructures section on the page. Wait for the configured status of the topology.

5. Connect to the application

Open info about the VM (VM IDs link in a table) and copy the public IP address.

Go to the following link:

https://paste_the_address:5904

Log in using your password and enjoy working with the Scipion.

Prepare Master node manually

Prerequisites (ubuntu packages)

• nvidia drivers
• docker with nVidia runtime
• X11 server running
• xserver-xorg xdm xauth nvidia-container-toolkit nvidia-container-runtime nvidia-docker2

Headless machines

Configure xdm

When running on headless machine (or a machine where nobody is playing FPS games all the time), make sure the X server accepts unauthenticated local connections even when a user session is not running. E.g., the /etc/X11/xdm/xdm-config file should contain:

DisplayManager*authorize:       false

However, such settings can be dangerous if the machine is not dedicated for this purpos, check for possible side effects.

Configure xorg

1. Run nvidia-xconfig --query-gpu-info to obtain the bus ID of the GPU. Example:

Number of GPUs: 1

GPU #0:
  Name      : GeForce RTX 2080 SUPER
  UUID      : GPU-4fcfbe08-eee6-df4b-59aa-4c867e089b2f
  PCI BusID : PCI:10:0:0

  Number of Display Devices: 0

2. Create an appropriate xorg.conf file for headless operation:

sudo nvidia-xconfig -a --allow-empty-initial-configuration --use-display-device=None \
--virtual=1920x1200 --busid {busid}

Replace {busid} with the bus ID you obtained in Step 1. Leave out --use-display-device=None if the GPU is headless, i.e. if it has no display outputs.

3. If you are using version 440.xx or later of the nVidia proprietary driver, then edit /etc/X11/xorg.conf and add

Option "HardDPMS" "false"

under the Device or Screen section.

Installation of prerequisites

nVidia runtime

Installation

https://github.com/NVIDIA/nvidia-docker/wiki/Installation-(Native-GPU-Support)

Configuration

You should now have the nvidia runtime installed, which is required to run this project.

You can now replace the default docker runtime runc by nvidia runtime.
Backup and edit file "/etc/docker/daemon.json".

{
    "runtimes": {
        "nvidia": {
            "path": "/usr/bin/nvidia-container-runtime",
            "runtimeArgs": []
        }
    },
    "default-runtime": "nvidia"
}

https://github.com/NVIDIA/nvidia-docker/wiki/Advanced-topics#default-runtime

If you need runc as a default runtime for some purpose, do not change this runtime. Note that you will now need to start the docker image with "--runtime=nvidia".

Run a container

Re-pull the images before running the containers

docker pull registry.gitlab.ics.muni.cz:443/eosc-synergy/scipion-docker:latest

Run

docker run --privileged -d --rm -p 5904:5904 -e USE_DISPLAY="4" -e ROOT_PASS="abc" -e USER_PASS="abc" -v /tmp/.X11-unix/X0:/tmp/.X11-unix/X0 --mount src=scipion-plugins,dst=/opt/scipion/software/em registry.gitlab.ics.muni.cz:443/eosc-synergy/scipion-docker:latest

Env var "USE_DISPLAY" will create new display (e.g. ":5"). Please note that you need new one for each instance. Therefore change the "USE_DISPLAY" value for each instance.

Only one-digit display number is now supported.

This is also related to the port. Change last digit of the ports "-p 5905:5905".

In addition, if you are using default docker runtime, you have to run the container with "--runtime=nvidia" parameter.

Basic test

Your instance should be available on the link: "https://your-adress:5905".

The log-in password is now hard-coded in Dockerfile.

Open terminal a try "nvidia-smi" and "glxgears -info" commands. Both should print output containing information about your nVidia graphics card.