Docker is a set of platform as a service (PaaS) products that use OS-level virtualization to deliver software in packages called containers. – Wikipedia: Docker
I have started an AWS machine to build Dockerfiles. The advantage is that the necessary Docker images are downloaded amazingly 🤩 fast on AWS. The other advantage is that the whole thing can run without blocking my working machine. Unfortunately, I had selected a default machine with only 8 GByte. Writing a Dockerfile which first runs a build process and then the proper image can fill up the disk quickly. Therefore, I had to cleaned up everything before I started the next build. If I forget it, then the build started and stopped after a certain amount of time with “no space left”. Originally this solution should make things faster and not cause additional work.
One advantage of cloud computing is that you can simply mount an additional disk into the system. But this alone does not help, since Docker still uses the directory from the 8 GByte disk.
Move the Docker directory
Of course, you will find a solution for this on Stackoverflow:
systemctl stop docker
mv /var/lib/docker /newvolume/docker-data
ln -s /newvolume/docker-data /var/lib/docker
systemctl start docker
- stop Docker
- move the Docker directory to the new hard disk
- create a link
- start Docker again
After that, docker images should show all images. But since Docker was stopped, the respective images have to be restarted.
Mounting an EBS volume
These are the commands to mount an EBS volume with EC2. I have copied them from How to Attach and Mount an EBS volume to EC2 Linux Instance as a backup:
lsblk
sudo file -s /dev/xvdf
sudo mkfs -t ext4 /dev/xvdf
sudo mkdir /newvolume
sudo mount /dev/xvdf /newvolume/
cd /newvolume
df -h .
… and don’t forget to auto mount the volume in /etc/fstab:
/dev/xvdf /newvolume ext4 defaults,nofail 0 0
Summary
Don’t worry if you need more disk space. Just create another volume, mount it and move the docker directory.
