This is post three in a development diary I am writing on setting up a new homelab in my home office. In the last post I described setting up the Linux VM and its Docker software to manage docker- based containers. Now it was time to load some containers onto the system.

The first few parts of the container install went very well. With Portainer installed, it was just a matter of selecting items on Docker Hub and then picking out the items I needed from the command line examples to load the containers.

• Heimdall
• SQL Server Developer Edition 2017

For SQL Server, I used 2017-latest. It would have helped if I got the port mapping right the first time though. I will file that in the list of things to check later on. Also, UFW was getting turned on at startup which did not allow access to Docker containers. I turned it off for the moment, but I will need to fix that in the future.

Autostarting

I am only auto starting Heimdall (which should serve as the entry point into my homelab) and, of course, Portainer. I’ll manually start SQL Server when I need it.

The problem

Aside from mapping the wrong port, everything went fine. Until I needed to upload the 20GB database we use for testing. Fox a user that is rather rusty on Linux and a novice with Docker, that was a little more difficult. Also, the sharing service I use claims to have FTP access, but I could not make it work. Here is what I finally did:

• Download via Windows to my desktop machine.
• Transfer the file via scp to the VM running docker (using the instructions at Linuxize). I used the transfer directory to hold the file.
• Install unzip and then decompress the file.
• Used docker cp to move the file into /var/backups on the container
• Used SSMS on my Windows machine to restore the backup using the default file locations.

Once this was done I did see a massive amount of memory being used on the VM. A restart of the VM fixed the issue.

Thumbnail by Delesign Graphics on IconScout