Even though I work in the field of cloud hosting and have learned a lot there, my blog was still down for a stupid error.
Out of curiosity I checked my analytics tool (umami) and could not reach it. It is running on the same server this blog is running on, so I had a look. I also checked my mails, since I did setup an E-Mail notification from UptimeRobot but since I do not get notified for E-Mails and do not check them that often I did not notice the downtime.
As you can see in the image I had some time in the afternoon to fix the problem.
Turn out my OS drive was 100% full ️?
I am using watchtower to automatically update my docker images and therefore the software on this server.
Every time when Ghost is pushing a new image to the tag ":4" watchtower will pull the new image and run the new version. That way I am always on the latest version.
But docker keeps all the images.
Create a systemd service and timer to delete unused images every day.
Here is the docker prune service:
cat /etc/systemd/system/docker-prune.service [Unit] Description=Delete not used docker images Wants=docker.service [Service] Type=oneshot ExecStart=/usr/bin/docker image prune -f [Install] WantedBy=multi-user.target
Here the docker prune timer:
cat /etc/systemd/system/docker-prune.timer [Unit] Description=Timer to delete not used docker images Requires=docker-prune.service [Timer] Unit=docker-prune.service # Every day at 4am OnCalendar=*-*-* 4:00:00 [Install] WantedBy=timers.target
That way I can keep the automatic updates and enjoy a less maintainance work for my hobbies like this blog and will not have any downtime because of this reason (hopefully).
I am surry for the around 10 to 20 people who visit my blog everyday, but honestly it will not kill me. I rather have a downtime (could be shorter next time ?️, I admit that) than old and unsecure software.
Thanks for reading, I wish you a great day without downtimes!