Docker guide

This guide aims to give some insights about running Docker containers, volumes and networks directly using docker run.

🚫 This page does not aims to be a comprehensive Docker tutorial, see the official documentation if you want to know more about Docker.

Basic Docker commands#

Run a container#

A Docker image can be run to create a container in 2 ways

-it Interactive: the process is attached to the terminal and you can stop it using ctrl+C
-d Detached: the process is detached and will continue if you exit the terminal.

Stop a container#

To get a Docker container ID and stop it:

docker ps
docker stop my-container

Automatically delete a container after run You can use the --rm flag to automatically delete a container after its run. It avoid accumulating containers and wasting memory.

docker run -it --rm my-container

Share volumes#

A lot of images will require input files, the best way to provide them is to share a volume from your system to the container. It can be done by using -v when docker run.

Let's say I have a /data/d2s-workspace/my-file.txt on my system, and I want to provide it as a container input file.

I need to provide he -v flag with paths separated by a :
- before the user provide the path of the shared volume on his system
- after the user provide the path it will be accessible from in the Docker container

docker run -it -v /path/on/my/computer:/path/in/container my-container \
    --input-file "/path/in/container/my-file.txt"

In this example we are sharing the /data/d2s-workspace volume from our system to /data in the Docker container. my-file.txt is then accessible as /data/my-file.txt in the container

docker run -it -v /data/d2s-workspace:/data my-container \
    --input-file "/data/my-file.txt"

Windows requires to provide the drive when sharing volume (e.g. c:) and the command to be one line.

docker run -it -v c:/data/d2s-workspace:/data my-container --input-file "/data/my-file.txt"

We usually provide a relative path using $pwd)

docker run -it -v $(pwd)/workspace:/data my-container --input-file "/data/my-file.txt"

To be tested on Windows.

Link and network#

Containers can be linked using 2 ways: --link flag and --network.

Network#

Especially used when running services using docker-compose. The --link flag method can be easier to use in simple cases.

Example of a network setup in the docker-compose.yml #

version: "3"
services:
  drill:
    build: ./apache-drill
    volumes:
      - /data:/data:ro
    ports:
      - 8047:8047
      - 31011:31010
    tty: true
    networks:
      - network
  
  graphdb:
    build: ./graphdb
    volumes:
      - /data/graphdb:/opt/graphdb/home
      - /data/graphdb-import:/root/graphdb-import
    ports:
      - 7200:7200
    networks:
      - network
  
networks:
  network:
    driver: bridge

Example of AutoR2RML connecting to the Drill container through a network#

docker run -it --rm -v /data:/data \
  --network network-pipeline_data2services \
  autor2rml -r \
  -j "jdbc:drill:drillbit=data2services-pipeline_drill_1:31011" \
  -o "/data/d2s-workspace/mapping.trig" -d "/data/d2s-workspace" -g "http://data2services/graph/autor2rml"

Link#

--link is deprecated, Docker recommends to use --network. But the --link flag can be more convenient to use in simple case: e.g. when linking to a single service, you just need to provide the container name instead of setting up a whole network.

How to use link#

Let's say you have a frontend application that needs to query a backend application.

Run the backend

docker run -it --rm --name backend -p 8080:8080 backend-image

To access it as http://localhost:8080 in the frontend

docker run -it --rm --name frontend --link backend:localhost frontend-image

To access it as http://backend:8080 in the frontend

docker run -it --rm --name frontend --link backend:backend frontend-image

You can link to multiple backend in a frontend

docker run -it --rm --name frontend --link backend-sql:backend-sql --link backend-redis:backend-redis frontend-image

Example of AutoR2RML connecting to Drill container using link#

# Run Apache Drill service
docker run -dit --rm -p 8047:8047 -p 31011:31010 --name drill -v /data:/data:ro apache-drill

# Run AutoR2RML, linking to the Apache Drill container
docker run -it --rm --link drill:drill -v /data:/data autor2rml \
    -j "jdbc:drill:drillbit=drill:31010" -r \
        -o "/data/d2s-workspace/mapping.trig" -d "/data/d2s-workspace" -g "http://data2services/graph/autor2rml"