docker-aem-base

AEM Base Docker Image

This is the base Docker image for AEM.

Depends on

Vagrant, VirtualBox, Docker

Also, install the vagrant-gatling-rsync plugin vagrant plugin install vagrant-gatling-rsync

Get Started

Start Vagrant which will get your VM (virtualbox) going: vagrant up

Create a directory called .sources and place in it a cq6.jar file and license.properties file.

If the docker command throws a TCP error, you probably need to setup the DOCKER_HOST.

export DOCKER_HOST_IP=$(vagrant ssh-config | sed -n "s/[ ]*HostName[ ]*//gp")
export DOCKER_HOST="tcp://${DOCKER_HOST_IP}:2375"

Then create the base image by executing the following:

docker build --tag="bhibma/aem-base" .

Next, build the author and publish images off the base image:

cd aem-author
docker build --tag="bhibma/aem-author" .
cd ../aem-publish
docker build --tag="bhibma/aem-publish" .

Then to run the author and publish images, execute:

docker run -i -d -p 4503:4503 --name publish bhibma/aem-publish && docker run -i -d -p 4502:4502 --name author --link publish:publish bhibma/aem-author

At this point you have two containers running - one for author and one for publish - and both are starting up AEM for the first time. This will be very slow. To speed things up on subsequent startups, we need to commit the running instance. We will first tag the base image so we can start a container from it if we want in the future:

docker tag bhibma/aem-author bhibma/aem-author:base
docker tag bhibma/aem-publish bhibma/aem-publish:base

Then next we will commit the running container as the 'latest' tag on the image so that when we start using only 'bhibma/aem-author' or 'bhibma/aem-publish' without any tag specified, it will run the running image, not the base non-running image.

docker commit -m "Running Instance" author bhibma/aem-author
docker commit -m "Running Instance" publish bhibma/aem-publish

Once that has been run, it should take <1 minute to start up the AEM instance from this these running containers.