Hi have a little ongoing discussion at work.
We're going to configure a server where the uptime is highly critical. A downtime will affect thousands of users. We will be configuring redundancy for the services that this server provides.
I'm for setting up the most hassle free solution with Debian. I'm considering installing Docker from Debian repos and not using the official Docker repos.
The plan is to install unattended upgrades, that installs security updates automatically. We'll do a manual apt upgrade when necessary. This will probably be once a month, where we do apt upgrade, reboot if necessary and check if everything is working.
We do this on our critical systems allready.
We're going to work with a partner that has a different approach at this. They like having the latest packages. They don't recommend unattended-upgrades. They recommend running Ubuntu Server and checking the changelogs for the critical services (in this case Docker).
They argue against our approach, since a distro upgrade might break things. In my eyes, we'll have the redundancy, which will give me the time to fix the unexpected issues.
I don't like their approach, since I think it is a higher risk. The changelog won't cover bugs and the process will be more time consuming. We have a lot of tasks and I'd like to work as efficient as possible, without getting any unexpected issues.
I'd like to hear your opinion. What's your approach and experience?
If I'm wrong in my strategy, I'd like to be corrected![😂]()
.
Thanks.
We're going to configure a server where the uptime is highly critical. A downtime will affect thousands of users. We will be configuring redundancy for the services that this server provides.
I'm for setting up the most hassle free solution with Debian. I'm considering installing Docker from Debian repos and not using the official Docker repos.
The plan is to install unattended upgrades, that installs security updates automatically. We'll do a manual apt upgrade when necessary. This will probably be once a month, where we do apt upgrade, reboot if necessary and check if everything is working.
We do this on our critical systems allready.
We're going to work with a partner that has a different approach at this. They like having the latest packages. They don't recommend unattended-upgrades. They recommend running Ubuntu Server and checking the changelogs for the critical services (in this case Docker).
They argue against our approach, since a distro upgrade might break things. In my eyes, we'll have the redundancy, which will give me the time to fix the unexpected issues.
I don't like their approach, since I think it is a higher risk. The changelog won't cover bugs and the process will be more time consuming. We have a lot of tasks and I'd like to work as efficient as possible, without getting any unexpected issues.
I'd like to hear your opinion. What's your approach and experience?
If I'm wrong in my strategy, I'd like to be corrected
.Thanks.
Statistics: Posted by Swampebob — 2025-02-07 08:31