The problem is more with zfs on consumer grade NVMes. I have/had problems in that configuration due to the bigger sector sizes. Proxmox itself does do frequent writes, but I don’t know how often exactly. I know that my problems went away with not using zfs.
https://www.reddit.com/r/Proxmox/comments/idlqh3/zfs_extremely_high_ssd_wearout_seemingly_random/
I wouldn’t suggest usb or sd-cards with proxmox due to its constant logging. You will fry them really quick unfortunately. Had that problem with NVMes.
For litterly anything else I would also suggest SD-Cards.
Hey, i have had the same trouble on an DL380 G9. Those bioses don’t support booting from PCIe at all. My server can’t even boot from drives from the Raid controller in IT-Mode.
I would suggest, by proxmox being a hypervisor, to just install proxmox on a single SATA disk and try to boot from there. This is what I have done in the end.
You can then use your NVMes as storage pool. Also you bifurcation can always also be a problem when trying to boot from those devices.
I would also as a last call try to disable bifurcation and see if one drive will show up. Maybe then you could use 2 real PCIe slots with cheap m2 to PCIe adapters.
Thanks for pointing out, corrected it right away.
Tbh all modern mainstream distros are lightweight I give you that. But there are always exceptions. Something like PopOS (I know not a server distro) can hog a lot of resources, so those are not suited.
Yes that’s correct. But I see 18 months maintaine windows for a complete distro upgrade is fairly often. Ubuntu Interim is in my opinion not really suited for server applications due to the small support windows.
Rocky Linux9 security EOL is in 8 years for the other end. In that context fedora is a lot more “short lived”.
You will need a pretty light distro since you only have 2GB of Ram. Normally I would recommend containerized workloads, but 2GB RAM are just a bit too small.
Your distro choice should also be made based on the frequency of maintainance and package availability.
In the server space you have some contenders.
Release based distros: Ubuntu is your beginner friendly go to recommended distro. Very well documented and with automatic security updates. In my opinion its okay but a tad bloated. Ubuntu has yearly release cycles but the LTS versions have longer support so you don’t have to upgrade your whole distro. Ubuntu uses apt package management.
Debian would be the next normal choice. Also apt based with almost yearly releases. No bloat, but also no auto features. You are more on your own. Similar to Ubuntu.
Fedora server is also a more beginner friendly got it all distro with better modularity and very recent packaging. Fedora uses dnf. Be aware that fedora has tight release cycles on which you have to upgrade every time. Fedora has virtually only a small grace period between releases.
Centos/AlmaLinux/RockyLinux are all RedHat Linux clones without the enterprise support but with the same packages. Rock solid distro used in the enterprise server industry. Very well documented and known. Due to enterprise world also a bit outdated. But I found packages that are newer here than in the Debian repos. Those distros also use dnf/yum.
OpenSuse Leap is also a Good distro. I can’t say much to it because I didn’t use it so much. Opensuse is well known and has a good knowledge base. There is also opensuse Tumbleweed wich is a rolling release distribution.
Rolling releases: Rolling releases are distros wich don’t have real release cycles but are more or less “rolling” no big upgrades needed but more of a once a mont maintenance type distro.
There is centos, archlinux, nixos, opensuse Leap and probably a lot more. Nixos is pretty special and I don’t really recommend it so much for beginners.
Last category auto updating, immutable micro distros wich are mostly used for container hosts. This distros are made for only hosting containers. You have to take care of the right storage setup and be aware of all the special quirks it comes with. Best ones are Fedora CoreOS, Flarcar Linux and Opensuse MicroOS. Those are “low maintaince” but only if you really know what you are doing. Steep learning curve and non standard procedures.
Hope this helps a bit.
Feel free to correct me :)
I suspect nextcloud having performance issues with slow Disk IO. With rootless containers I had a much worse performance than rootfull. Also using MySQL Backend instead of SQLite did speedup the performance.
Nevertheless I have the same problems with nextcloud as you stated. Pretty much not as usable as I thought.
Well because of money. You certainly have to pay to get Ubuntu certificated. And you only do this to have a Linux system with support from the manufacturer.
It’s an enterprise problem with an enterprise solution.
The normal personal systems are not in the same segment.