#Docker Files for Various Projects

The goal of this repository is to provide docker images that host a complete environement for development related projects and also to promote and quickly show the capability of Butor framework.

---

##Currently Available

• java : creates a basic Ubuntu 14.04 image with Oracle JDK 7
• mule : based on codingtony/java and installs a vanilla version of Mule ESB 3.5.0
• tomcat7 : based on codingtony/tomcat7 and installs a vanilla version of Tomcat 7.0.54
• rhodecode : the source management platform for Git and Mercurial repository (v. 1.7.2)
• artifactory : A popular opensource repository manager for Maven
• jenkins : A popular opensource continuous integration (CI) engine
• cdh5-base : A base image with Cloudera Hadoop 5 repositories + Oracle JDK 7. Based on Ubuntu 12.04 (precise)

##Planned

• haproxy : Haproxy 1.5
• mysql : MySQL (or MariaDB) 5.5
• butor-demo webapp
• butor-demo services
• butor-demo data

Quick Tutorial on Docker persistance

If you need to persist the data to a volume, you will need to build a data only container and create volumes on it. Then you will need to use volumes from that data container when you start your other containers. Here's how to proceed :

Creating the data only container

Here I want the volumes /opt/rhodecode and /var/repo, and we need to name it so that we can refer to it later. I use tianon/true since it is probably the smallest Docker image available.

docker run -v /opt/rhodecode -v /var/repo   --name "rhodecodedata" tianon/true

Then you want to use the volume from that named container when you start your image. The changes you do in /opt/rhodecode and in /var/repo will be persisted. Initialy theses directory will be EMPTY

The way to do it :

 docker run -ti --volumes-from rhodecodedata codingtony/rhodecode bash

If you don't add the --volume-from rhodecodedata to the command line, you will see the original content from the image

 docker run -ti -p 5000:5000   codingtony/rhodecode ls /opt/rhodecode

When you are ready to start (as a daemon) an image with the change saved in the volumes you can simply do :

docker run -d -p 5000:5000  --volumes-from rhodecodedata codingtony/rhodecode

VERY IMPORTANT

The volume you try to mount in the image must not be symlinks, they must be real paths! Otherwise you will get "Cannot start container... not a directory" when you try to start the image with --volumes from

Backup a Docker Data Container

Automatically

I've built a tool to easily backup volumes from data containers

Fetch backupContainer.sh from this repo :

wget https://raw.githubusercontent.com/codingtony/docker/master/backupContainer.sh
chmod +x backupContainer.sh

And backup easily with :

./backupContainer.sh containerdataname /backup/destination/directory

This will create a timestamped tar.gz file named /backup/destination/directory/backup-containerdataname-YYYYMMDD_HHmmSS.tar.gz

Enjoy!

Manually

If you prefer to do it manually :

This create a timestamped tgz file in the current directory containing the /var/repo and /opt/rhodecode from your data image

docker run  --rm --volumes-from  rhodecodedata busybox tar cpf - /var/repo /opt/rhodecode | gzip > backup-rhodecodedata-$(date +"%Y%m%d_%H%M%S").tar.gz