You must log in or # to comment.
A hopefully “success in progress”: I am building a new trueNAS server for storage. I have a k8s cluster and am currently using rancher for storage, but I decided at my scale central storage made more sense & would be easier to manage. I am also using that opportunity to upgrade from 2TB usable storage to 44TB usable storage. Fingers crossed everything will work 🤞
I finally set up a VPN instead of exposing unnecessary ports to the wild!
Nothing broke
I cannot claim the same.
The truest answer :)
I’ve been self-hosting for years, but with a recent move comes a recent opportunity to do my network a bit differently. I’m now running a capable OpenWRT router, and support for AdGuard Home is practically built into OpenWRT. I just needed to configure it right and set it up, but the documentation was comprehensive enough.
For years I had kept a Debian VM for Pi-Hole running. I kept it ultra lean with a cloud kernel and 3 gb of disk space and 160MB of RAM, just so it could control its own network stack. And I’d set devices to manually use its IP address to be covered. AGH seems to be about the same exact thing as Pi-Hole. With my new setup the entire network is covered automatically without having to configure any device. And yes, I know I could’ve done the same before by forwarding the DNS lookups to the Pi-Hole, but I was always afraid it would cause a problem for me and I’d need an easy way to back out of the adblocking. Subjectively, over about 6 years, I only had a couple worthless websites that blocked me out.
I haven’t yet gotten to the point where I’m trying to also to intercept hardcoded DNS lookups, but soon… It’s not urgent for me because I don’t have sinister devices that do that.
Building to this week. A few months ago, I was given a broken nas. I took it, thinking I’ve at least got 16TB of storage if it won’t work. Fixed it. Saw the software includes docker, and then saw it has just 2GB ram and before I installed anything it would complain about low memory. Got 16GB, and installed it last weekend.
Spent the week installing and setting up Immich, navidrome, and integrating my other server running arrs.
16gb. So you got a $300 nas
Pulled 8 out of an old laptop, and bought 8 for about 30€.
All of my apps are running without issue. First time in months
I’ll go first: I got XMPP (Prosody) setup for the family.
Also, less this week (cheating a little), but I’ve setup all my services with SSL (self-hosted root CA), domain names, and (finally) a dashboard (Heimdall.)Edit: I can’t sepll.
Nice, had my XMPP server now running for a couple of weeks, not many users on it so far though. But my highlight of the week was managing to get the slidge whatsapp bridge with Prosody running, so I at least don’t have to use the official app anymore for all those people who resist to get off of it.
Nice, same! Was also really positively surprised by how great the Android app(s) for XMPP feel.
Only thing not working yet for me is group chat creation. Oh well. Maybe this weekend.
On the other hand though, voice and video calls have worked flawlessly.
Nice, same! Was also really positively surprised by how great the Android app(s) for XMPP feel.
We’re on iOS and I wish I could say the same. Looking at the Android apps makes me very jealous.
Only thing not working yet for me is group chat creation. Oh well. Maybe this weekend.
What server software are you using? I went with Prosody and it felt pretty easy to setup the muc module for groups, but, on the other hand, I haven’t gotten around to voice and video calls.
Are you using sturn/turn server? Almost always needed for calls and video, you should join prosody support channel that are really helpful xmpp:prosody@conference.prosody.im?join
No, not yet, that’s why I haven’t set it up yet. Hopefully its a this-week thing.
Ah, too bad. IMO better clients would make it drastically easier to convince people to switch.
Hm, I can create groups (also with muc), and the other members are added, but writing a message triggers “x left the group” for everyone. Dunno. Probably something trivial I overlooked. But honestly… Weather is too good today to be bothered 😄
Ah, I already had a TURN/STUN coturn server set up for matrix and jitsi, so it was just a matter of telling prosody about that. So I cheated a little I guess 😄 Here is my full config for that, in the unlikely event that you’re using NixOS.
I had enough time to install sort of pihole.
I’m curious what alternative to Pi-hole you set up. (I’m planning on installing Pi-hole soon but wanna hear all my options)
I have used Adguard Home before. I found it to be very similar to Pi-hole.
I never tried Technitium.
Currently I use Pi-hole with unbound.
Acronyms, initialisms, abbreviations, contractions, and other phrases which expand to something larger, that I’ve seen in this thread:
Fewer Letters More Letters AP WiFi Access Point CA (SSL) Certificate Authority DNS Domain Name Service/System Git Popular version control system, primarily for code IMAP Internet Message Access Protocol for email IP Internet Protocol SMTP Simple Mail Transfer Protocol SSL Secure Sockets Layer, for transparent encryption TLS Transport Layer Security, supersedes SSL VPN Virtual Private Network VPS Virtual Private Server (opposed to shared hosting) XMPP Extensible Messaging and Presence Protocol (‘Jabber’) for open instant messaging k8s Kubernetes container management package
[Thread #142 for this comm, first seen 7th Mar 2026, 06:40] [FAQ] [Full list] [Contact] [Source code]
good bot
I plugged in an NVIDIA gpu in my server and enabled ollama to use it, diligently updated my public wiki about it and now enjoying real time gpt: OSS model responses!
I was amazed, time cut from 3-8 minutes down to seconds. I have a Intel Core7 with 48gb ram, but even an oldish gpu beats the crap out of it.
In that same vein I got an AMD Pro V620 32GB off ebay and have been struggling to get it to POST on my x570 motherboard, but I finally tried it on my old ASUS b450-i with a Ryzen 5 2400GE and with a few BIOS setting changes it fired right up.
Now I need to figure out what I’m doing wrong on the x570 board so I can run the V620 combined with my 9060XT for bigger models
What GPU and model you use?
NVIDIA Corporation GA104GL [RTX A4000] (rev a1)
From lspci
It has 16gb of VRAM, not too much but enough to run gpt:OSS 20b and a few other models pretty nice.
I noticed that it’s better to stick to a single model, I imagine that unload and reload the model in VRAM takes time.
I finally got around to installing Jellyfin. Still trying to get hardware transcoding working. I think I have it set up, but it still wants to use the CPU. I’m thinking permissions but I ran out of time.
Fun project.
I think QSV is the new “easiest” way if you have an Intel CPU. Here are some docker compose values that might help:
group_add: - "110" - "44" devices: - /dev/dri/renderD128:/dev/dri/renderD128110isrender44isvideoYou can
grep render /etc/groupto find your values.I found CPU accelerated transcoding to be as effective as using GPU acceleration for my small media server setup. Nvidia wasn’t worth it for me.
Oh thanks! I didn’t have the group_add.
Why the group add? Does JF default user not have access to dev dri?
The table (dm) might finally make the switch from roll20 to foundry for a campaign!
I got a test box set up with nixos and a config that runs all of my services. I wanted to test the declarative rebuild promise of it, so I:
- Filled the services with my some of my backed up data (a copy of the data, not the actual backup)
- Ran it for a few days using some of the services
- Backed up the data of the nixos test server, as well as the nixos config
- Reinstalled nixos on the test box, brought in the config, and rebuilt it.
And it worked!!! All serviced came back with the data, all configuration was correct.
I’m going to keep testing, and depending on how that goes I may switch my prod server and nas to nixos.
Very cool!
Re: the backup / restore of state in NixOS: I found myself writing the same things over and over again for each VM/service, so finally wrote this wrapper module (in action e.g. here for Jellyfin), which confgures both the backup services and timers, as well as adding a simple
rsync-restore-jellyfincommand to the system packages. In case you find this useful and don’t already have your own abstractions, or a sufficiently different use case 😄This is great! Thanks
This week I saw my 3 machine cluster flailing trying to stay online, digging around identified it as an issue with communication with my NAS. It was running NFS3 and so I swapped that to NFS4.1 and did some tuning and now my services have never been faster!
I’m redoing everything I have from scratch. This week I have FreeIPA set up from OpenTofu + Ansible configs, and enrolls most of my other servers against FreeIPA. I am still migrating TrueNAS to use FreeIPA’s Kerberos Realm for auth, and I need to chown a lot of files for the new UIDs and GIDs homed in FreeIPA. After that, I’m setting up FreeRadius for auth to switches, APs, and Wifi. And then after that, I’m back to overhauling my k8s stack. I have Talos VMs running but didn’t finish patching in Cilium. And after the real fun begins.
