

Check lsmod and see if the kernel modules are loaded. Nvidia smi is probably loading them for you on first run.
Infrastructure nerd, gamer, and Lemmy.ca maintainer


Check lsmod and see if the kernel modules are loaded. Nvidia smi is probably loading them for you on first run.
Edit your bootloader config and turn off quiet / splash so you actually get a useful boot log.
Imagine a disk shelf of 24 drives being connected by a pair of sas loops. You’d want the faster speeds then. It’s not about individual drives.
One is 6gb sas and the other is 12gb.
You probably won’t notice a difference


Holiday sale starts Dec 18th


Doesn’t really make sense for self hosting. Filtering the traffic is pointless if the traffic just completely overwhelms your internet.
Also generally self hosters aren’t running bgp with their own asn.
You’ve made a virtual disk on the zfs. The vm will never see the zfs, that’s managed entirely by the host.
Yes you’ll want to make a normal partition inside that virtual disk.
With vms you can’t just access the host zfs, it’s always abstracted. If you use lxc containers on proxmox then you can bind the zfs into the container (google it for steps, it’s not in the Gui)


I believe notifications would be accessible. Note that i don’t mean the push notification backend mentioned by Doomerang, but the actual notification that goes into your status bar (which is all processed on device). That would be readable by the OS in theory.
“i did it on accident” blows my mind. It’s by accident, not on accident.
Yes. It’d be pretty silly of it not to.
I just gave it my giant lightroom library.
I used to run it on about 800 prod boxes and we provided the hosting for the Gentoo forums. It’ll always have a soft spot in my heart, but even with us using binary packages I won’t miss how long emerge takes.


I don’t think any lemmy instances can afford to run on the big clouds.
FWIW I did this with jellyfin and ended up just using a vm instead of lxc. This way I could just pass the entire device through, not have to mess with drivers in my proxmox host, and not have to reboot all my vms/lxc just to apply updates.
The ceo stepped down and now its just being rolled fully into Microsoft.
https://www.axios.com/2025/08/11/github-ceo-dohmke-step-down
Fractal design, node case. One of them is a cube that takes 8x 3.5 and microatx. I used one for years and it was great.


Is your spine straight?


Are you not in any group chats with 50 people trying to organize something? They become a cluster fuck.
For direct chats, totally agreed. I’d be seriously insulted if you used an ai to talk in a 1:1 conversation.
Observium is an easy starting point for network gear, without going down the Zabbix rabbit hole.
Personally I just roll Victoriametrics and Grafana.
At my house around 10-15. For lemmy.ca and our other sites, 35ish maybe. At work… hundreds.