Shoutout to transactional updates

Just a quick shoutout to the amazing transactional server role of openSUSE (Leap). I’m running most of my servers as transactional servers since more than a year, and I had so far zero issues with the automated update procedure. This is great!

I know that all of my servers are always up-to-date because they manage themselves. This has been a huge automation improvement for me - more free time for us. Before I had a monster ansible playbook, which was nice, but always required manual intervention.
Now I know that also when I’m gone for a multi-day hiking tour, all of my servers are constantly updated because they will install the updates themselves (and roll back in case of issues).

I still have two NAS systems under my control that I want to update manually because of zfs. As zfs is not part of the openSUSE QA procedure I want them (and only them) to not do stuff automatically yet. After more than a year without major issues I however start to warm up to the idea of also switching them to transactional servers.


In short: I’m running almost all of my server as transactional server in one or the other form for more than a year and I had zero issues so far by doing so. For me this concept in conjunction with the strong openQA testing procedures that SUSE and openSUSE has has turned out to be a great server system. It’s always fresh, always up-to-date also when I’m not around to update it manually.

This is great.