Portability is the key for me, because I tend to switch things around a lot. Containers generally isolate the persistent data from the runtime really well.

Docker is not the only, or even the best way IMO to run containers. If I was providing services for customers, I would definetly build most container images daily in some automated way. Well, I do it already for quite a few.

The mess is only a mess if you don’t really understand what you’re doing, same goes for traditional services.

Most likely, a Hetzner storage box is going to be so slow you will regret it. I would just bite the bullet and upgrade the storage on Contabo.

Storage in the cloud is expensive, there’s just no way around it.

There was a good blog post about the real cost of storage, but I can’t find it now.

The gist was that to store 1TB of data somewhat reliably, you probably need at least:

• mirrored main storage 2TB
• frequent/local backup space, also at least mirrored disks 2TB + more if using a versioned backup system
• remote / cold storage backup space about the same as the frequent backups

Which amounts to something like 6TB of disk for 1TB of actual data. In real life you’d probably use some other level of RAID, at least for larger amounts so it’s perhaps not as harsh, and compression can reduce the required backup space too.

I have around 130G of data in Nextcloud, and the off-site borg repo for it is about 180G. Then there’s local backups on a mirrored HDD, with the ZFS snapshots that are not yet pruned that’s maybe 200G of raw disk space. So 130G becomes 510G in my setup.

Imagine if all the people who prefer systemd would write posts like this as often as the opposition. Just use what you like, there are plenty of distros to choose from.

• Open the GUI network settings
• Set DNS to the IP of the PiHole, make sure the “automatic” switch is off.
• Do the above for each active interface (ethernet, wlan) and for both IPv4 and IPv6
• Save/apply settings
• Turn the interface(s) off, then back on
• resolvectl flush-caches just in case

Look at resolvectl dns to check there’s no DHCP-acquired DNS servers set anymore

If you use a VPN, those often set their own DNS servers too, remember to check it as well.

This is true, with a couple gigs of RAM and SATA storage Nextcloud is not at all bad. Assuming an instance with not that much simultaneous users.

It feels like slow sometimes, then after an hour with M365 at work it doesn’t feel slow at all.

There’s a base image of ublue, which is Silverblue without a DE. I’d suppose you can mostly just layer e.g. Sway or i3 on top.

Traditional package model will still have it’s usage, of course, I agree. But if Silverblue works for a developer like me, I’d say a for more “regular” users immutable distros seem like a very viable option.

I recently put the nvidia variant of ublue-os on my work laptop, which has Optimus graphics. Couldn’t be happier.

It’s great to see these variants popping up! I really think ostree may be the future for desktop Linux, and not even very far away.

Thank you!

Rsyslog to collect logs to a single server, then lnav for viewing them on that server is a good combo. Oldschool but very effective for self-host scale.

Glad the tip was useful!

For a bit enhanced log file viewing, you could use something like lnav, I think it’s packaged for most distributions.

Cockpit can be useful for journald, but personally I think GUI stuff is a bit clunky for logs.

Grep, awk and sed are powerful tools, even with only basic knowledge of them. Vim in readonly mode is actually quite effective for single files too.

For aggregating multiple servers’ logs good ol’ rsyslog is good, but not simple to set up. There are tutorials online.

Devuan is more stable

So Devuan has even older versions of packages than Debian? Stability in the distro context means that features, APIs, UIs don’t change. Please don’t mix software bugs with stability.

It may be I’ve entirely misunderstood how systemd works, but I think your description of it is off by a mile too.

but a different init starts a new process ID for each separate program

Of course there are PIDs with systemd too! First of all, systemd itself has a PID (1).

For systemd, which runs system wide to handle everything, if one program locks, systemd has to make adjusts for the whole system to fix the problem.

This is just wrong… Sure, if the service in question is dependent on a lot of other services, or vice versa. If your programs tend to lock, that’s the application’s fault and should be handled at the application level.

I found Artix to run smoother or lighter than Arch.

This is most definetly a difference in what else is running on the system. Systemd doesn’t really use that much resources. Unless you are measuring RAM usage in the megabytes. Which is of course valid on constrained systems, but on a regular desktop one browser tab will need orders of magnitude more resources than any init system.

I want Firefox running an isolated process from the one that Plasama desktop is running

This just shows you have absolutely no clue on Linux processes, I really really doubt anyone is running Firefox under systemd. And neither have you.

There are valid reasons for choosing a different init system, but you have not provided a single one that is really true. It seems like you are only repeating things heard from some one else.

The difference is systemd is one thing to handle everything

This is true, but it refers to systemd handling a lot more than process management. Systemd has the problem that nowadays it does log management, memory management, login management, user management etc. This goes against the UNIX philosophy of one tool for one job, and THAT is why people frown on systemd.

I’d second this. Fedora is great, don’t get me wrong, but it’s not rolling or stable.

I think stable was referring to not crashing here.

Most Debian based distros, actually.

So Chrome and Google VPN? /s