Docker Container for Anaconda 5.2.0

This is a Docker image based on osixia/light-baseimage for running Anaconda 5.2.0. It start jupyter and 'cron' as services (which will restart after quiting or a crash).

With over 6 million users, the open source Anaconda Distribution is the fastest and easiest way to do Python and R data science and machine learning on Linux, Windows, and Mac OS X. It's the industry standard for developing, testing, and training on a single machine Anaconda.

Usage (minimal)

docker run -it --name=anaconda \ 
              -v <path for data files>:/mnt/data:rw \
              -e DGID=<gid>
              -e DUID=<uid> \
              -e TZ=<timezone> \
              -p 8888:8888 \
              logicwar/anaconda

Start/Stop

You can then start/stop the server with :

• docker start anaconda or docker start -i anaconda if you want direct interaction
• docker stop anaconda

Shell access

For shell access while the container is running do :

docker exec -it anaconda /bin/bash

Server interaction

For server interaction you can attach by using :

docker attach anaconda and then Control-p followed by Control-q to detach

From an attached console press Control-c to stop the server.

Data access

If you are not using a 'persistent' volume but the auto-created volume, then you can determine it by using :

docker inspect -f '{{ .Mounts }}' anaconda

It will return something like this :

[{volume 6f58390b223b8427a8515ea4aad9279e4ed618d2e2eb1d425e0d6f98d88fec6c /var/lib/docker/volumes/6f58390b223b8427a8515ea4aad9279e4ed618d2e2eb1d425e0d6f98d88fec6c/_data /mnt/data local  true }]

You will then need root rights to access the /var/lib/docker/volumes/6f5839...8fec6c/_data folder.

Parameters

• -p 8888 - jupyter port
• -v /mnt/data - local path for the installation
• -e DGID for GroupID - see below for explanation
• -e DUID for UserID - see below for explanation
• -e TZ for timezone information : Europe/London, Europe/Zurich, America/New_York, ... (List of TZ)

User / Group ID

For security reasons and to avoid permissions issues with data volumes (-v flags), you may want to create a specific "docker" user with proper right accesses on your persistant folders. To find your user uid and gid you can use the id <user> command as below:

  $ id <dockeruser>
    uid=1001(dockeruser) gid=1001(dockergroup) groups=1001(dockergroup)

and finally specify your "docker" user DUID and group DGID. In this exemple DUID=1001 and DGID=1001.

Logging in and setting a jupyter password

• Login with the token (which can be found in the running console) and set a password (choice at the bottom of the login page) on : http://your_host:8888
• Stop/restart the container to take into account the password (or just quit jupyter from interface as it will restart automatically)

Setup a cron job to run a python script

With Conda environments you activate them by typing source activate your_env_name and deactivate them by typing source deactivate. Unfortunately this is not how you set it up to run via cron. You will need to find the path to the python that runs when you are in that particular environment.

To avoid running the script as root, you should use the docker user.

As root in a bash terminal to your container, run the following commands:

  $ (crontab -u docker -l ; echo "* * * * * /home/docker/.conda/envs/<your_env_name>/bin/python <your_python_script.py>") | crontab -u docker -
  $ service cron reload

The above command will run the script every minutes.

 * * * * * *
 | | | | | | 
 | | | | | +-- Year              (range: 1900-3000)
 | | | | +---- Day of the Week   (range: 1-7, 1 standing for Monday)
 | | | +------ Month of the Year (range: 1-12)
 | | +-------- Day of the Month  (range: 1-31)
 | +---------- Hour              (range: 0-23)
 +------------ Minute            (range: 0-59)

To remove your job:

  $ crontab -u docker -l | grep -v '<your_python_script.py>'  | crontab -u docker -
  $ service cron reload

Sample of a simple docker-compose.yml

version: "3"

services:
  anaconda1:
    image: logicwar/anaconda:latest
    volumes:
      - "/MyContainerPersistentMounts/Anaconda/Data:/mnt/data:rw"
    ports:
      - "8888:8888"
    environment:
      DUID: "1004"
      DGID: "1003"
      TZ: "Europe/Zurich"
    container_name: "anaconda-1"
    hostname: "anaconda-1"
    network_mode: "bridge"
    tty: true
    stdin_open: true

Versions

• V0.1 Initial Release