I have a single-vm deployment I am playing with using bosh (following these instructions: http://concourse-ci.org/concourse-lite.html)
After a bunch of builds, my pipeline errors with
/scratch/docker/tmp/GetImageBlob128904326: no space left on device
and on my mac, if I go into ~/Library/Containers/com.docker.docker/Data/com.docker.driver.amd64-linux
the Docker.qcow2 has pretty much grown to my max ~64gig
My pipeline is located at https://github.com/buildit/twig/blob/concourse/concourse/pipeline.yml
The only thing I can think of is that for my e2e tests, I am copying the docker files I use to spin up a docker-compose environment into each build so it doesn't go grab them by itself (see https://github.com/buildit/twig/blob/concourse/concourse/scripts/run_e2e_tests.sh) and perhaps those e2e environments are not being cleaned when the step fails (I am failing on purpose right now)
Is there a step I need to take to clean up these containers? I have 27 containers when I run fly -t lite volumes