Provides several full Docker environments for running XWiki.

The following configurations are currently supported:

• Two Docker containers with one container for running latest MySQL 5.x database (configured to use UTF8 and be case-insensitive) and another container for running the latest Tomcat 8 + Java 8 + XWiki (the version depends on the branch/tag you use).

All source files are under the LGPL 2.1 license.

Assumptions

The goal is to provide a production-ready XWiki system running in Docker. This why:

• The OS is based on Debian and not on some smaller-footprint distribution like Alpine
• Several containers are used with Docker Compose: one for the DB and another for XWiki + Servlet container. This allows the ability to run them on different machines for example.

Using

You should first install Docker on your machine.

Then there are several options:

1. Get the sources of this project and build them.
2. Just pull the xwiki image from DockerHub.

Building

This is the simplest solution and the one recommended. Here are the steps:

• Install Git and run git clone https://github.com/xwiki-contrib/docker-xwiki.git or download the sources from the GitHub UI. Then choose the branch or tag that you wish to use:
  • The masterbranch will get you the latest released version of XWiki
  • The 8.x branch will get you the latest released version of XWiki for the 8.x cycle
  • The 8.4.4 tag will get you exactly XWiki 8.4.4.
  • etc.
• Go the directory corresponding to the configuration you wish to build, for example: cd xwiki-mysql-tomcat.
• Run docker-compose up
• Start a browser and point it to http://localhost:8080

Note that if you want to set a custom version of XWiki you can checkout master and edit the env file and set the values you need in there. It's also possible to override them on the command line with docker-compose run -e "XWIKI_VERSION=8.4.4".

Note that docker-compose up will automatically build the XWiki image on the first run. If you need to rebuild it you can issue docker-compose up --build. You can also build the image with docker build . -t xwiki-mysql-tomcat:latest for example.

Pulling existing image

This is a bit more complex since you need to have 2 docker containers running: one for XWiki and one for the database.

Here's a minimal Docker Compose file using MySQL that you could use as an example (full example here:

version: '2'
services:
  web:
    image: "xwiki/xwiki-mysql-tomcat:latest"
    depends_on:
      - db
    ports:
      - "8080:8080"
    volumes:
      - xwiki-data:/var/lib/xwiki
    environment:
      - MYSQL_USER=xwiki
      - MYSQL_PASSWORD=xwiki
      - MYSQL_DATABASE=xwiki
  db:
    image: "mysql:5"
    volumes:
      - ./mysql/xwiki.cnf:/etc/mysql/conf.d/xwiki.cnf
      - mysql-data:/var/lib/mysql
    environment:
      - MYSQL_ROOT_PASSWORD=xwiki
      - MYSQL_USER=xwiki
      - MYSQL_PASSWORD=xwiki
      - MYSQL_DATABASE=xwiki
volumes:
  mysql-data: {}
  xwiki-data: {}

Details for xwiki-mysql-tomcat

Volumes:

• Two volumes are created:
  • A volume named <prefix>_mysql-data that contains the database data.
  • A volume named <prefix>_xwiki-data that contains XWiki's permanent directory.
• To find out where those volumes are located on your local host machine you can inspect them with docker volume inspect <volume name>. To find the volume name, you can list all volumes with docker volume ls.
• Note that on Mac OSX, Docker runs inside the xhyve VM and thus the paths you get when inspecting the volumes are relative to this. Thus, you need to get into that VM if you need to access the volume data.

MySQL:

• To issue some mysql commands:
• Find the container id with docker ps
• Execute bash in the mysql container: docker exec -it <containerid> bash -l
• Once inside the mysql container execute the mysql command: mysql --user=xwiki --password=xwiki

Future

• Setup xinit
• Configure libreoffice
• Solr as external service

Credits

• Created by Vincent Massol
• Contributions from Ludovic Dubost, Jean Simard
• Some code was copied from https://github.com/ThomasSteinbach/docker_xwiki. Thank you Thomas Steinbach