Hi
I have fiddled with linux since 1993, but have never had linux as my primary job. I have managed a few servers (10-15 servers) in a couple of my previous jobs. I have patched each of these servers manually. In november '20 I started a new job as a linux admin with responsibility for 300+ virtual linux servers. I have set up Ansible/AWX and written playbooks that works great for patching, deployment and other day to day tasks. I have scheduled monthly patching of all systems and it works quite well. On patch tuesdays I just lean back and wait for patching to finish. I have patched everything on all of the servers.
Recently I got responsibility for for an additional 100+ servers and the owners of those systems wants to apply security patches only. I have googled and asked a question on linuxquestions.org and it seems there's no reason to not grant their wish to only apply security updates.
However, in my earlier jobs, I have "inherited" (RedHat/CentOS) systems that have never been patched and they have broken badly when I have tried to patch them due to broken dependencies and bad (no) management from the previos sysadmins. Luckily they were virtual servers, so I was able to revert to snapshots. Today my servers are about 30% CentOS/RedHat and the rest on Ubuntu LTS and we are in the process on migrating everything to Ubuntu LTS.
So to my question: If I in the lifetime of servers, which might be up to 5 years (LTS), I need to apply all updates on servers that has only gotten security patches, is there a risk of getting into dependency hell? Or any risk of other breakage?
My gut says that we should full patch systems regularly to prevent broken dependencies and other potential problems. Is my gut wrong?
Bookmarks