social.dk-libre.fr is a Fediverse instance that uses the ActivityPub protocol. In other words, users at this host can communicate with people that use software like Mastodon, Pleroma, Friendica, etc. all around the world.
This server runs the snac software and there is no automatic sign-up process.
RE: https://mastodon.uno/@giuseppegv/115730518668624174
I like the idea of this post. Here is mine with some additions of my own:
OS
- Desktop and laptop
- #Debian 13 with #Cinnamon
#homelab server
- #Proxmox 9 with #Debian and #HomeAssistant OS VM's
Email
Service providers
- @disroot and @Tutanota
Client
- #Betterbird
Notes
- @logseq
Browser
Desktop
- @mozilla #Firefox , @librewolf , and @torproject
Mobile
- #IronFox
VPN
- @mullvadnet
Password and TOTP Manager
- @keepassxc with #Syncthing
Search
- #DuckDuckGo , @ecosia , and @marginalia
Social
- @Mastodon and @pixelfed
#FOSS #Debian #GNULinux #GNU #Linux #Proxmox #homelab #homeassistant #disroot #tuta #tutanota #betterbird #logseq #Mozilla #Firefox #Librewolf #Tor #Torproject #IronFox #Mullvad #MullvadVPN #VPN #TOTP #KeePassXC #Syncthing #DuckDuckGo #Marginalia #Mastodon #Pixelfed
My router interface traffic graphs are pretty useless. Every night around 2:00am, various things run backups. Some do backups across the network, others copy out to an off-site location. But the bottom line is that if I look at my network interface graphs, that massive spike every night ruins the Y axis.
These are the daily, weekly, and monthly graphs. And the fact that daily hits around 450M, but ordinary hours are like 8-10M, it means I can't see the rest of the graph. It means that I can't spot changes or interesting things.
I'm not really sure how to do better. (It's amateur hour over here)
Diese Woche geht's bei @ajfriesen und mir um Monitoring im Smart Home und dem Homelab. Welche Arten von Monitoring gibt es? Wieso ist das wichtig? Welcher Ansatz eignet sich fürs Smart Home? Darüber hinaus berichtet Andrej über seine Migration von Fritz zu UniFi für WLAN, LAN und Router.
Im Podcatcher eurer Wahl oder hier: https://podcast.smarthuette.de/episodes/grune-balken-machen-glucklich-monitoring-im-smart-home
Viel Spaß beim Hören!
#homeassistant #homelab #unifi #monitoring #uptime_kuma #zabbix #prometheus #nagios #icinga #grafana #linux #docker #traefik
A new 40kg #UPS for my #HomeLab =)
Actually, this is to power critical home infrastructure during power outages. A gasoline generator is good, but a charging station is quite cheap to charge (compared to gasoline). Also, they can work together for UNLIMITED POWER*.
*UNLIMITED POWER means while I have gasoline in canisters. It’s around 40 liters. It’s around 60 hours of power from the generator. Plus 10 hours from #Bluetti, so the amount of power is quite finite and would end in three days.
Today was the last move. It is all in one rack. It’s all halfway decent. I had a major downtime though because I had a stupid #xcpng goof that required upgrading a server that had live workloads on it. I couldn’t migrate them off. That sucked.
But it looks ok now. And it is so much more compact. I am using some air filter fabric over the front door to filter the air. And I have the front facing inside and the back blowing out toward the outside. I figure the air is cleaner on the house side of the garage.
So, I have been using Zitadel as my IDP so far, but it’s far too clunky and slow to my taste
I’ve been thinking about moving to Authentik, but now I see this project, PocketID, and I’m starting to think it would do the job at hand very easily
Only this is it’s only supporting passkeys, but I guess Bitwarden will take care of that
Convince me not to use PocketID, how has the experience been with this around here ?
#homelab #selfhosted #selfhosting #pocketid #authentik #zitadel #idp #oidc #security #passkeys
Le temps investi dans la prévention (mise à jour) est toujours infiniment plus court que le temps passé à résoudre une compromission. La sécurité et la stabilité ont un prix : celui de la veille technologique.
Point final.
Remembering my homelab before I gained experience with containers.
The thrill of running multiple PHP websites on the same box hoping the next system update won't break any of them 🙄
...or natively installed Jellyfin, Pi-Hole.
Fragile configurations.
Not anymore 😊
Yes, it is an abstraction layer, but no: It's not a black box.
You can inspect the containers, enter them, take a look at the scripts they run, read their logs, take a look at the Dockerfile they were built from.
It's a great way for maintainers to ship their project, because the image is identical and easy to debug, without having to ask the users which distro, libraries, etc. they are using while filing a bug.
So: Take a look at docker. It's worth learning.
I have a small golang app that listens to #zfs zpool events and sends a notification to ntfy.sh or your self-hosted instance of it when there’s state changes with errors. I wonder if more people would be interested in using it. I currently have it running in both #freebsd and #linux servers. #homelab
| Yes, I would use it on my servers: | 13 |
| No, I have no use for it: | 0 |
Self-Host Weekly #148: Maintenance Mode
#MinIO alternatives, software updates and launches, a spotlight on #Poznote - a lightweight note-taking app, and more in this week's #selfhosted recap!
https://selfh.st/weekly/2025-12-05/
#selfhost #selfhosting #opensource #foss #homelab #homeserver #devops #sysadmin #privacy #security #homeautomation #app #apps #software #newsletter #fediverse
Just published one of my favorite pieces I get to write each year - my favorite #selfhosted apps launched in 2025.
Let me know what I missed!
https://selfh.st/post/2025-favorite-new-apps/
#selfhost #selfhosting #opensource #foss #homelab #homeserver #development #devops #sysadmin #server
Self-Host Weekly #149: A Few of My Favorite Things
My #favorite apps of 2025, software updates and launches, a spotlight on #wastebin -- a lightweight #pastebin app, and more in this week's #selfhosted recap!
https://selfh.st/weekly/2025-12-12/
#selfhost #selfhosting #foss #opensource #homelab #newsletter #devops #sysadmin #development #privacy #security #fediverse #software
After it was deemed not to be good enough for Windows 11 by Microsoft, I salvaged my dad's old computer (2nd gen Intel Core i3) and turned it into a Linux server.
I am considering using it as a lab for my pet projects and experiments, replacing a ~4 EUR/month Hetzner server (cx22) in exchange for a comparable CPU power and similar cost in electricity bill and plenty more RAM and storage.
I wrote a blog post about running #coreos in #incus in my #homelab.
bout image import, using ignition and persitent data management.
https://agnomi.net/posts/incus/fedora-coreos/
#selfhosted
It took me almost a month to actually get these #Aqara THA sensors delivered. Had 2 orders cancelled because the item was unavailable…
Also, I remember buying this kind of sensor for 5$ back in the day…
This way, you can simply use the built-in rules via the interface, but still use enhanced rules via ProxLB by tags or pools (even for non HA VMs/CTs). This feature was already requested a long time ago and get's now streamlined over all ways, ensuring we can balance our VMs (and CTs) across the nodes with and even without any configured HA objects - fully flexible to your needs!
Requirements:
* Proxmox 9.x
* ProxLB 1.1.11
Version 1.1.11 is expected to be released by the end of this or within the upcoming week; depending on my spare time.
PS: Keep in mind, tomorrow starts our "Virtualization Gathering" at @credativde@mastodon.social and I'm providing a talk about my @BoxyBSD@bsd.cafe project where I tell you a bit more about the infrastrucutre/hypervisors (due to the limited time only very high level)! Feel free to join, it's free!
Tags: #opensource #PVE #homelab #enterprise #vmware #virtualization #cloud #onprem #kvm #pve #resourcescheduler #loadbalancer #loadbalancing #drs
@nicholasburns Most of it is SAS. But where local storage bit me was in migrating from one host to another as I upgraded XCP 8.2.1 to 8.3. When you have shared storage, moving a node from one compute to another is trivial. When I have only local storage, migrating VMs has been really difficult. It's either time-consuming (20-120 minutes to move a big VM) or even impossible (version-to-version migration on XCP has been difficult with only local storage).
I'm going to admit that I am doing something immature in my #homelab and I'm looking for opinions. I've got multiple #XCPng hosts, all using local storage. I have no NFS or iSCSI storage. That's kinda silly. Shared storage is super useful and I'm literally not using it.
Unless I go to some serious effort to make a high-performance SAN, I expect network storage performance to be so-so for VM storage, but maybe I'm too pessimistic. I currently only have copper gigabit in the rack. No fiber, no 2.5G copper or anything like that. I'm not sure if that's going to be viable for NFS or iSCSI.
I could dedicate a host to running TrueNAS Core with a bunch of storage. But what has always bugged me about this is that my storage host becomes a single point of failure for all the compute nodes. #TrueNAS is super reliable but everything has to reboot once in a while, and these stupid enterprise-grade servers take anywhere from 4-8 minutes to boot. If I had a single storage node, and I needed to reboot it for an OS upgrade, everything would hang for a while. That's no good. Not updating the OS on the storage system is also not good.
So what am I supposed to be doing for shared storage on a #Xen cluster? How do I avoid a storage host becoming a single point of failure? How do you update and reboot a storage node, without disrupting everything that depends on it?
sysop
boostedI think #FOSDEM might be a perfectly good slot to provide people BSD based systems for free again.
#opensource #learning #education #runbsd #freebsd #openbsd #netbsd #bsd #freevm #freevps #hosting #ipv6 #homelab
This is artisanal code, typed by hand and with love (and a wee bit of swearing and restoring backups) by me, taking inspiration from a lot of tips and links, shared here in the #HomeLab group. I couldn't have done it without y'all, thank you for sharing :) No "AI" was involved in creating this shell script.
Logo source: https://www.aimeecozza.com/human-made/
3/3
A very dangerous and brutal shell script using `exiv2` or, alternatively, when no EXIF data found, the last modification timestamp, to sort pictures into neat directories called YYYY/MM. Set the `odir=` at the top to your preferred output directory.
WARNING: Do not just copy/paste and let it run on your precious picture archive. This is raw, aggressive code that doesn't care about collateral damage. You have been warned.
https://codeberg.org/jwildeboer/gists/src/branch/main/2025/20251207PictureEXIFTips.md
1/3
We're coming even closer to you again with our new locations which will be available soon for deployments:
* UK, London
* Norway, Sandefjord
* Sweden, Stockholm
* Canada, Toronto
* US, Kansas
#BSD #RUNBSD #FreeBSD #NetBSD #OpenBSD #VM #VPS #FreeVPS #FreeVM #education #opensource #homelab #learning #ipv6 @gyptazy
Project Rebel Homebase – Teil 2 ist live. Erinnert ihr euch an Teil 1?
Jetzt wird’s physisch: Statt VM-Kuschelwelt gibt’s echtes Metall, echte Hitze und sechs Raspberry-CM-Blades, die nur dank Airflow nicht sofort in die Knie gehen.
Der Unterbau steht – das Skelett der Rebellion.
https://www.pandolin.io/project-rebel-homebase-teil-2-das-skelett-der-rebellion/
#RebelHomebase #Homelab #Linux #OpenSource #FOSS #Kubernetes #k3s #DigitaleSouveränität #CloudRebelAlliance
I think I'm gonna form a new band called "Cardinality Explosion". And the title album will be "All the metrics I shouldn't be exporting". Presented to you by "Random Operator Records". 💥
Hm. Does anyone have a little shell script that does this:
- Find all picture files
- Check if it has EXIF data and if yes, rename the file to YYYYMMDD_HHMMSS using the EXIF creation date
- If NOT, rename the file to the same pattern, but use the filesystem last modified date
I have `for img in $(ls *.[Jj][Pp][Gg] 2> /dev/null); do exiv2 -r'%Y%m%d_%H%M%S_'"$(tmp=${img%%.*};echo ${tmp##*_})" rename "$img" ; done`whicj does the first part, but fails when no EXIF
I upgraded #xcpng and I just noticed that it has some IPMI integration. It's able to tell a few things about power and temperature and display it at the bottom. It just discovered all these attributes. I didn't do anything to turn that on.
And while doing sysadmin stuff, I also updated all 3 #forgejo runners to the current 12.1.0 version. Now to write that blog entry on how to setup Samba on a Linux box that allows you to create and use Time Machine backups for Apple machines. I learned a lot about that topic in the past few days :)
It's Friday, so as always I did the `dnf update`, followed by a `reboot` on my little pool of 8 servers. I also deployed a certificate with certbot on my new NAS box, using certbot and my own certificate authority. For details on that: https://jan.wildeboer.net/2025/07/letsencrypt-homelab-stepca/ The weekend can start :)
I started kicking the tires on n8n this week. The appeal for me is that I can run it on my #homelab and it was a one-click installation via Coolify.
Even a few hours into playing with it, I can already see a few subscriptions I was paying for won't be needed anymore.
I had this recent exchange with my wife:
Wife: What is FreeBSD?
Me: It's Unix, an operating system, and it's awesome!
Wife: Oh OK. The only reason I asked is because I am finding these FreeBSD CDs around the house. FreeBSD 12.0, 13.3, 14.0.
Me: Yeah, I use the old versions as coasters for my tea mugs.
Wife: *pauses, laughs* I can see that.
#FreeBSD #homelab #tech #fun #unix #computers #freebsd15 #operatingsystem #awesome
It’s coming together. I have 2 out of 6 racked. All the power is in the rack. Network and all the actual workloads are still on the bread rack behind and to the right. I will get these 2 servers online, move a bunch of workloads to them, and then I’ll be able to move more servers off the bread rack to the new rack.
I don’t see any way to redo the networking without some brief outages. I’ll have to disconnect the router, move it, and reconnect. A few minutes of disruption.
That’s going to be the hardest part: the main network wire runs through the whole basement and pops out in the garage and it’s basically at its limit. I have about 5-6 feet in the garage and that’s it. So when the main router finally moves to the rack, the rack has to be in its final location and then it can’t really move much.
Either that, or I just stick one of these RJ45 couplers on there and extend the line. I only have 200Mbit service, so I’m not worried about losing bandwidth to a coupler.
So I bought a full-sized rack for the garage. Might be a bit too big to call it a #homelab any more 😜
It came with 2 PDUs that take 3-phase, 60A, 250V AC. I, um, won’t be plugging those into the house. My servers run on 240V, 2-phase. This plug is just a monster. My hand in the photo for scale.
I’m tickled that it says “Sun” on the side. I got my start in #sysadmin work on Sun IPCs and Sparc 4s back in the day.
At this weekend I finally established another network in my home — a ZigBee network. A looong time ago (in 2010 year) I touched the ZigBee networking in my university (ITMO, previously IFMO, in Saint-Petersburg) — these times it was a new technology, not used widely. And as a student I have some fun time playing with ZigBee main router, supplemental router and end-devices. You can view old photos and screenshots of old software on a my extremely old blog: http://h0rr0rr-drag0n.blogspot.com/2010/04/cc2530zdk.html (and read a blogpost, if you understand Russian).
It is kindly fascinating, that now, after 15 years, I can just buy some ZigBee-powered devices from AliExpress (using Black Friday discounts) and connect them to the network inside my house right in the way I did it in the university 15 years ago!
Sadly, although I bought native supported main router device, based on the EFR32MG2 with some software from Ember (EZSP v8) inside, the OpenHAB doesn't support this device natively — it supports it, but since my server is running NetBSD, I got problems with some bundled with OpenHAB things. Looks like some native libraries (rxtx-java) don't have bundled NetBSD versions. And the same library in the repository built for Java 8, not for Java 17.
So, I decided to use Zigbee2MQTT, not to build the necessary Java library myself. It was kinda scary — use program, which connects my ZigBee network via ZigBee USB-dongle to the MQTT server — which is written on JavaScript
. Not on the C (as I can totally understand, for a such low-level program, operating with embedded devices) or at least on the C++/Perl/Python/whatever. But, looks like it works good enough, if I don't try to pair the device in wrong mode (my window sensors has two modes to pair them with network: first "common" mode causes zigbee2mqtt to silently crash and the second "compatible" mode works without problems).
And I could understand now, why people has so much problems with smart home security. Installed MQTT server mosquitto — it allows unauthenticated connections by default. Installed zigbee2mqtt — it allow connections to frontend without any password by default
At least these two services don't each much memory: 1.2 Mb for Mosquitto and 75.6 Mb for ZigBee2MQTT.
For now, my ZigBee sensors works pretty well and robust, like these devices from university 15 years ago 
I picked up a Home Assistant Green to replace my broken RPi3 Home Assistant box last holiday season. After a year of using it, it's been stable and an excellent addition to my #homelab
We had to replace our washing machine this year, and it has Home Assistant integration, which is pretty cool. I haven't quite figured out what to do with this data, but I'm debating flashing something on our TV or changing one of the bulbs' colors so that I see it.
(28 Nov) Self-Hosting is Rising and Linux Users are Leading This Revolution
Self‑hosting isn’t anti‑cloud; it’s pro‑agency. It’s choosing the right locus of control for the things you care about.
https://itsfoss.com/self-hosting-rising/
Archive: ia: https://s.faithcollapsing.com/ltae5
How do you guys block YouTube shorts in your network? I'm not sure if Adguard Home or RouterOS can do this.
Digitale Souveränität fängt im Kleinen an. Statt nur abstrakte Cloud-Services zu konsumieren (Lego Duplo), baue ich mir meine Infrastruktur selbst (Fischertechnik).
Mit: Raspberry CM5, Uptime Blades, 3D-Druck & Lötkolben. Willkommen in der #RebelHomebase.
Im Blog erkläre ich das philosophische und technische „Warum“ hinter meinem Mini-Datacenter auf dem Schreibtisch.
#DigitaleSouveränität #FOSS #Homelab #Linux #OpenSource #Kubernetes #k3s #CloudRebelAlliance
New blog post!
It is about how this Mastodon instance is backed up daily without downtime =)
https://www.yevi.org/blog/daily-matrix-and-mastodon-backups-with-zero-downtime
#Mastodon #Matrix #SelfHosted #SelfHosting #SelfHost #HomeLab #Ansible #Proxmox #Backup #PBS #ProxmoxVE #Blog #GlitchySocial
Boostez votre #Homelab Proxmox ! 🚀
Fini les manipulations complexes de scripts. Cette interface web moderne (Next.js/React) vous permet de gérer les scripts communautaires PVE du bout des doigts. Un must-have ! ✨
➡️ Le Repo: https://github.com/community-scripts/ProxmoxVE-Local
#Proxmox #ProxmoxVE #PVE #Selfhosting #SysAdmin #DevOps #WebUI #OpenSource
Finally going through some of the backlog of things i've been wanting to do for my #homelab . Moving over to #komodo + #gitea + #renovate to manage my #docker containers. I have them split based on categories in proxmox lxc containers. I use 2 lxc containers, ha-containers, for important containers, and media-containers, for well media containers.
ha-containers is now fully on komodo, pinned to releases and updating my important containers is now easy peasy and arent falling behind updates.
#opensuse Leap 16.0 is out! I have some free time today so I'd figure I'd try it out in a VM.
Initial impressions:
New, streamlined installer just requires a few clicks with minimal input required. In a few minutes the installer finished. Rebooted and BAM! The OS boots straight into this mess!
This is not XFCE with Wayland. Time to troubleshoot.
...sigh...
The new hard drives for my NAS (Network Attached Storage) build have arrived. Two WD (Western Digital) RED 8TB. I will go shopping now after unpacking, so they can adapt to room temperature for a few hours before I install them and power them up. Never forget that step! Wait until they have reached room temperature. And then wait a few hours more. When you spin them up too early, they might get damaged. Don't ask why I know ;)
Upgraded my cute homelab [1] with a simple power meter that sits before the UPS. The whole thing, so 3x Lenovo Tiny PCs, a max 120W 10 port network switch with PoE, a Raspberry Pi 4B, the 4 port KVM switch (both PoE powered) and the 500VA UPS uses 50-55W when idle, up to 120-130W under load. I call that very reasonable :) 1/3
Silly idea for a computer home lab: you probably have a couple old smartphones lying around, gathering dust.
Maybe you can build a compute cluster from them though? They are probably some of the most efficient compute devices available. Also passively cooled so zero noise! With OTG, it might be possible to even connect to them serious storage and build a Ceph cluster...
https://hackaday.com/2025/04/09/self-hosting-a-cluster-on-old-phones/
#tinkering #homeserver #homelab #computers #sysadmin #upcycling
🚨 Alerte Proxmox 9 ! 🚨
Mes conteneurs Docker/LXC ne démarraient plus après la mise à jour containerd.io... 🤯
J'ai trouvé l'origine du problème (le conflit AppArmor/runc) et je vous explique la seule façon de le corriger (ou de l'éviter !) pour relancer vos services rapidement.
Le guide complet est sur mon wiki 👇 https://wiki.blablalinux.be/fr/proxmox-9-apparmor-docker-lxc-conflit-demarrage
👋 The End of an Era 👋
- Fact: time is an irreplaceable critical resource, too often in short supply
- Resources: finite patience, focus, attention; infinite idgaf
Requisite Changes
- Core hardware from HomeLab will be moved to the colo this month (systems which matter)
- Likely much will be given to specific OSS projects for free, easiest for those local to Milpitas (colo)
Respond to thread or DM if you want to be updated with the inventory spreadsheet (partially complete, more to add)
I'm in Australia right now. I can boot my home PC via Wake On LAN to access my files if I have to. I placed a webcam in front of it so that I can verify that it actually does so.
It's weird seeing my computer 16.500 km away from me. It feels like my own little Mars mission. Especially with how SSH feels at 300-400 ms latency.
And yes, that's a drinking glass I forgot on my desk.
The Self-Hosting Starter Pack: 5 Simple Tools I Recommend To Get Started With Your Homelab https://itsfoss.com/self-hosting-starting-projects/ #Homelab #List📋
I have had bad luck with HDDs lately.
I bought 4*16TB refurbished, one of them is broken.
One 2TB had checksum issues at my dad's house. I have given him a new HDD of the same size just in case. He replaced the drive. I replaced the drive in the pool. But the pool stays degraded. The drive has so many errors it has been disabled.
That can't be the drives. We have replaced at least 3 or 4 drives and we still have checksums issues. That must be the cables or the sata ports.
Just installed good old trusty `trac` https://trac.edgewall.org/on my system. I'm deploying with Ansible and on a VM, and it was failing while setting the locale. It needs `LC_ALL` set, which for some reason was not set in my system. Set it with `update-locale LC_ALL=en_US.UTF-8`.
Upgraded my Raspberry Pi 4B from Fedora 42 to Fedora 43 using the offline upgrade process. I documented what worked for me as a gist at https://codeberg.org/jwildeboer/gists/src/branch/main/2025/20251107UpgradeToFedora43Raspi4.md and couldn't have done it without the little tip from @krist — Thx, Krist!
Sur le blog, un rapide billet de réflexion sur le sujet #HomeLab, Cluster #Proxmox et #WakeOnLan https://blog.genma.fr/?HomeLab-Cluster-Proxmox-et-WakeOnLan
J'ai un #synology DS1515+ qui vient de cramer...
Me voilà avec 2 + 1 disques 4To ... (2 du NAS + 1 de sauvegarde hors toute électricité)
Qu'en feriez-vous à ma place ?
J'ai déjà dans mon #homelab 4 x 6To et 3 x 3To et je n'ai plus de place dans mon boîtier(les 3To ont quasi 90 000 heures au compteur... ça va lâcher bientôt !).
#NAS #ugreen ? J'ai envie de voir ... mais bon j'essaye de ne plus acheter par geekitude ... la planète n'a pas besoin de ça...
Boîtier usb3 ?
#Homelab 2022 Part 2 - #Samba on #SmartOS #Illumos using delegated datasets : https://thetooth.name/blog/homelab-2022-part-2-samba-on-smartos-using-delegated-datasets/
Pretty cool trick so you don’t have to deploy a fully fledged Anubis to protect against DDoS-by-LLM-companies?
Yesterday I completed the migration from Authentik to Pocket ID on all of my homelab services!
I also included Tinyauth to link OIDC to a service without a login page.
Today, there was a part of me that silently made me think that get rid of my secondary domain and buying a new one was a good idea.
So here I am. Melting on the sofa after 10 hours of Ansible, OIDC clients, creating users, tweaking config files, VPN, Jenkins security and jobs configurations, plus your regular system administration.
Oh, did I mention that it was supposed to be my day off? 😅
#homelab #selfhost #selfhosted #selfhosting #sysadmin #systemadmin #linux #linux #ansible #oidc #vpn #jenkins #sso
I'm setting up a kubernetes service running PS3netsrv to serve PS3 games from my network storage to the PS3 I just jailbroke. How's your halloween going?
In early September, The Matrix Foundation homeserver went down.
I'm extremely proud of our SRE team. They had a Disaster Recovery Plan and monthly exercises to apply it, resulting in no data loss despite a 24h outage.
I've learned a lot about how to properly backup/restore a Postgres database when writing this post with SREs. We also learned how to better prevent and be resilient to human error.
Thanks all for the hugops during the outage!
sysop
boostedThis release adds features that were requested during the #DutchProxmoxDay2025 which was hosted by our friends Tuxis (tuxis.nl). I'm hearing your feedback and all your feedback from the community and user base is important to me, so let me introduce the following new features:
* Add an optional memory balancing threshold
* Add affinity/anti-affinity support by pools
* Add pressure (PSI) based balancing for memory, cpu, disk
* Pressure (PSI) based balancing for nodes
* Pressure (PSI) based balancing for guests
You can simply upgrade via APT or by obtaining the latest Docker/Container image. PSI based balancing (still in beta) requires Proxmox VE 9 or greater.
Thanks to my employer
@credativ@mastodon.social (https://credativ.de) who provides me the required time to work at this.
GitHub: https://github.com/gyptazy/ProxLB
GitHub Release: https://github.com/gyptazy/ProxLB/releases/tag/v1.1.9
Author: @gyptazy@gyptazy.com
#PVE #Proxmox #Community #OpenSource #FOSS #gyptazy #DutchProxmoxDay #coding #projects #virtualization #vms #hypervisor #homelab #vmware #alternatives #pve #balancing
It's been a long day. Tomorrow though, I really need this MikroTik CCR2004-1G-2XS-PCIe, running the home router, to be quiet. Therefore, lay in bed, figure out which Thermal Grizzly conductive pad is most appropriate for replacement between the NIC's heatsink.. and maybe there a STL file for making a carbon fitment plus copper core (taken from an old Broadwell system, heatsink with a couple of blah blah blah... yep. Go to sleep brain, stop this, maybe more tomorrow.
So GPT-Pro and Deep Research... tell me what to buy.
#homelab #tired #mikrotik #intel #mellanox #networking #sleepy #uuuuuuugh #gptpro
Any #ZFS folks want to offer me an opinion on the following? I have 2 ZFS pool of raidz2 with 7 2TB SAS drives each (about 11.5T). I want to add a LOG drive. I have a single SAS SSD that's 300G. When I suggested using a 1TB LOG drive earlier, someone said that was way too much. (which is why I picked up a used 300G SSD)
My question is whether it makes sense (is possible) to partition that drive into 2 partitions and have one pool use 1 partition as a log drive (e.g., 150G) and the other pool use the other partition as a log drive on the same physical device.
Is this going to be worse because it's just too much IO on one device? Is it reasonable? Any other ideas?
🔐 Nouvelle automatisation Proxmox !
Marre des mises à jour manuelles ? 😩
Nouvelle publication sur le Wiki : Automatisez la mise à jour complète de tous vos Conteneurs LXC et VMs (Debian/Ubuntu) sur Proxmox VE grâce à deux scripts Cron.
➡️ Gagnez du temps, restez sécurisé.
Lien direct : https://wiki.blablalinux.be/fr/script-update-lxc-vm-proxmox
Disponible aussi sur ByteStash (solution de stockage d'extraits de code) : https://bytestash.blablalinux.be/s/a055d43f24f8b58218b4cbdad100b7ec
#Proxmox #HomeLab #SysAdmin #DevOps #Automatisation #Linux #cybersécurité
I have a CA in my homelab that gives out short lifespan TLS and SSH certs. The lifespan is 48 hours, and they renew daily on a cron (Their crons are randomly spread out throughout the day so they don't all renew at once). I have alerting rules to monitor for any with lifespan <24 hours, and fire alerts to let me know somethings wrong.
Spotted the problem? The hour went back last night, as BST ended. All my servers use local time, not UTC. Every single TLS and SSH cert had an hour today where their certs were <24 hours from expiring, but hadn't hit their renewal cront time yet.
My email and Discord alerts have not. stopped. pinging. all. day. (I have a *lot* of certificates in this lab lol) gonna have to fix this before next year!
My little MCU temp on a battery experiment was interesting. We have the house thermostat at 68 in the day and 65 at night. Why such a stairstep down at night with an 8 degree swing? How come there are these ~2 degree jumps this morning?
See, I left Facebook more than 10 years ago, my home has been Microsoft-free since around 1998. I run my #Homelab where I store all my digital stuff (photos, scanned documents, LibreOffice files etc) on a NAS (Network Attached Storage) server that runs Linux. I have never used AWS or Azure, if I need some compute capacity, I rent VPS (Virtual Private Server) at EU based hosting companies. I did #DigitalSovereignty long before the term even existed. You can become the owner of your data too!
1/3
I need some #Linux touchers and #HomeLab gurus to explain in excruciating detail what I'm doing wrong.
I want my #JellyFin server to be served over HTTPS.
I had nginxproxymanager running, but it has stopped working with Let's Encrypt.
What's the *simplest* way to add HTTPS to #JellyFin? It's an Ubuntu box with the default Jellyfin installed (no Docker or anything complicated like that).
Informed speculation welcomed!
I've been on the hunt for some accessible, easily self-hosted (on prem or datacenter) chat solution for a while now, and I've been looking very closely at deltachat/chatmail.
Not gonna lie, the fact that just uses mature SMTP facilities for transport is making me weirdly aroused.
If anyone has experience running their own chatmail server or just using deltachat in general, I'd be glad to hear from you.
It's a very bad day for #Minio users. (Ceph RGW is a mature alternative, though not a trivial migration for most.)
https://github.com/minio/minio/issues/21647
#ceph #opensource #homelab #kubernetes #containers #cloudnative #DevOps
I'm building a #Dell server and I have a couple choices of disk controller. The chassis is an R730xd.
My choices are:
The system will run XCP-ng on bare metal, and host a whole lot of general purpose BSD and Linux VMs and VPSes. It's got 320G of RAM and 14 2Tb SAS drives.
Thoughts? Boosts welcome. Suggestions for totally different hardware or software configurations not welcome.
If anybody is using #dell servers in their #homelab and you have super noisy fans, I just discovered this Docker container solution that will set the fans to a reasonable speed.
I'm using a 3rd-party PCI SAS controller, and that seems to cause the server to go apeshit. Fans sound like they're ready for liftoff. This fixes it right away.
#selfhosting #selfhosted
If you have an #ASUS router from the current or last gen and want to do some #HomeLab networking, then I recommend #MerlinWRT. If you're using AiMesh (whether or not as a mesh network), then only the router(s) need it. The ASUS stuff should continue to work (app, DDNS, etc.), but you'll have more control over the finer points of your network to do LAG, run your own DNS, etc., from the router.
https://www.asuswrt-merlin.net/
Things have been going will with the #HomeLab #Selfhosted instance; no crashes, recovers from failures, performs automated nightly backups etc.
Why wouldn’t I mess with that?! 😆
Is there a way to manually prefilter an #email account?
Get some cheap harddrives on eBay, they said (well I said...). You'll save money I said..
This is just dumb. I bought inexpensive used drives, but then for various reasons I didn't get around to even powering them up until well past the return window on eBay.
So when I finally got them spun up? 12 out of 14 completely useless.
#homelab #selfhosting #ebay
Configuring a new #Dell server in the #homelab. Who the fuck makes up settings like this!?
“Select timeout value. (0 defaults to 4 seconds, 15 is no delay, 1-14 is timeout value in seconds)”
0 is 4,
1 is 1,
2 is 2,
4 is 4,
15 is 0
#selfhosting #selfhosted
hm i am thinking of what distro to choose instead of #nixos for my server...(i want to migrate away since i am tired of all the drama)
really all i need is NVIDIA drivers, libvirtd and docker since my whole setup is already dockerized(and a forgejo runner in a vm), so really i can choose any distro with glibc.
J'ai installé uptime kuma qu'on m'avait conseillé ici, , et ça marche très. Je cherchais un truc aussi simple pour faire de la supervision système (disque, CPU, RAM, network), et je viens de découvrir #beszel https://beszel.dev/ 🤩
Un serveur qui centralise les données, des agents qui se déploient en ligne de commande fournie , un peu de conf SMTP et hop ! Les alarmes fonctionnent, les graphes pour les métriques basiques, tout mon #homelab supervisé en 30 minutes !
#autohebergement
Je suis en train de monter mon #homelab mais la question qui me tourmente, c'est la supervision.
#cacti ? #librenms ? #prometheus ?
Autre ?
Je cherche quelque chose de simple qui me permettent d'ajouter une machine et d'alerter sur des seuils d'alerte.
J'ai commence avec #prometheus, qui me semble tres bien, mais j'ai pas envie de passer du temps sur grafana pour creer un dashboard.
Des propositions ?
#autohebergement #supervision #homelab
My relay at https://fedi-relay.gyptazy.com has currently 139 instances connected, mostly tech related sharing the same mindset and interests like #Linux, #BSD, #Ansible, #Proxmox, #Coding, and many more! You can easily join from your instance when using #Pleroma, #snac (#snac2), #Mastodon and its forks 🙂
#fedi #fediworld #fedicommunity #community #FreeBSD #OpenBSD #NetBSD #homelab #Python #Debian #RockyLinux #Feditips
Migrated my ZigBee2MQTT to FreeBSD 14.3, of course, running inside a Jail (and with IPv6!) 🙂
All running perfectly fine and as stable as you'd expect from FreeBSD
Plus the rc.d script is so minimalistic and simple, that I don''t get it, why anyone would prefer intransparent systemd to that !
Another part of my home-infrastructure migrated over!
Great success! I've returned my instance back to local storage, all in aid of reducing costs in other cloud platforms.
I had set up nginx to cache my S3 media in SSD cache space, so now I need to modify that config to do the same for my media living on high-capacity spinning rust.
I know that as a small instance this isnt really needed, but who knows, maybe I'll have a sudden influx of new users and will be glad of having it set up!
😅 Synology Reverses Policy Banning Third-Party HDDs After NAS sales plummet
「 The policy, introduced earlier this year, made third-party HDDs from brands like Seagate and WD practically unusable in newer models such as the DS925+, DS1825+, and DS425+ 」
If you've a single server with a bunch of disks (see picture) would you rather:
* Create two independent LVMs 1 for the 4x 2TB disks and another one for the 20x931.5G disks? (with 1 luks volume per physical disk)
* Same as above but with 1 luks volume per logical volume served by the LVM.
* Glusterfs with encryption feature enabled.
* Glusterfs with 1 luks volume per physical disk.
* ...?
What would you go with and why?
C'est l'outil idéal pour enrichir votre bibliothèque personnelle en mode "coller l'URL et oublier les tracas" ! 💾
🚀 Installation Express (Merci la Communauté !)
Vous voulez l'essayer tout de suite ? Utilisez un docker-compose.yml préconfiguré :
🔗 Lien direct pour le Docker Compose : https://bytestash.blablalinux.be/s/7649ba8ffd903c2bb061877405dcddbf
Hashtags Populaires
#HomeTube #VideoDownloader #HomeLab #Plex #Jellyfin #SelfHosted #Docker #Proxmox #SponsorBlock #NoAds #AutoOrganization 🎬🚀
tootctl media remove --days=7ETA: 41:56:22
Ah. This'll take longer than just a single coffee break. Ah well. See you all on the other side of a great media purge!
After that, it's syncing my B2 bucket back to my local disk!
I want planning on building a new server, but I couldn't resist a deal on an absolute unit of a case with space for at least 13 hard drives, so I guess I am. #homelab
I’m self hosting things, and if I don’t write things down I don’t remember how to do them. So far I’ve been relying on Obsidian for my docs, but I’m really tempted to go with @atuin desktop.
No BS playbooks you can run (so playbooks you keep updated), and that you can store on git? Open source and free? No AI?! It’s almost too good to be true!
I use mattermost as a free, open source Slack replacement. I like it a lot. I also use authentik as my SSO provider for my #homelab / #selfhosted stuff.
If you don't pay for the enterprise version of Mattermost, you don't get the SSO features. You can fake that by using their free GitLab integration. But you also don't get the capability to switch users from one form of authentication (email/password) to SSO after you set it up.
This is pretty obscure, and it involves a bunch of raw Postgres queries. But I figured it out and wrote a blog on converting Mattermost users from email/password to SSO
@rachel I'm interested in what you end up doing. I have exactly the same setup and do exactly the same thing.
I use portainer to run all my containers. So when I read your description, I thought "does portainer have an API?" it does.
There's a /stack/{id}/git/redeploy API that (if you know which stack ID you just modified?) you could invoke. #Portainer would re-pull and redeploy the stack. So in my case I might be able to make a job that runs on commits and invokes a couple APIs to get the stack ID and then tell it to pull and redeploy?
I'd have to change what I do so that it pulled from some specific branch (like release or something) or always do my development on a branch and have it pull/deploy from main. That seems do-able but I've never done it.
Ok I have a setup I'm trying to get a better handle on. Currently I have a setup at two edge locations running a few similar docker containers handling tls termination plus few containers unique per node They are managed manually by copying docker-compose files via git push/pull. Not exactly ideal, I know. Options I've considered: I think I'm somehow leaning towards k3s? Or independent single-node Talos instances?
* Leave as is, I don't need to touch these often enough to bother to automate
* Ansible? I'd have to spin back up and I'm not using it anywhere else, bleh
* Terraform? Also bleh, no good provider that I see, would have to re-write everything in HCL? Bleeeeeeeh.
* Same as I have been doing but with podman/systemd/rootless? Maybe idk
* K3S single node clusters on each, pull them into ArgoCD?
* Talos nodes at each edge location and join them to my main cluster over wireguard? Probably not worth the downsides?
New blog post! I just got a Framework Desktop, and I wanted to use it as a node on a Talos Linux Kubernetes cluster. The initial boot was not easy to grasp, and took me a few hours to figure it out, so I wanted to share with you how to make it work. I hope this is useful for someone!
@homelab @siderolabs @frameworkcomputer
#HomeLab #TalosLinux #Framework #FrameworkDesktop #Kubernetes #Blog
https://mteixeira.wordpress.com/2025/09/28/booting-talos-linux-on-a-framework-desktop/
This year, I overhauled my entire home/lab network, ditching proprietary, locked-down gear for open-source freedom! Goodbye, Ubiquiti Dream Machine and Unifi APs.
Hello, barebones x86 PC with Intel Core i7-8550U and 6 Intel Gigabit Ethernet interfaces, running pure FreeBSD 14.3-RELEASE. Connects via PPP, handles traffic with pf firewall, and rocks an IPv6-first setup. Robust, reliable, and it "just works"!
It's powerful enough, that I can run some bhyve VMs and small application Jails alongside without an issue.
There's a #bug in the #NextCloud updater, where it complains about an extra file found called `REUSE.toml`. According to https://github.com/nextcloud/server/issues/55111 it's safe to remove the file by hand and retry the upgrade.
Is anyone out there hosting Open Cloud? From what I gathered, it's a fork of ownCloud Infinite Scale, that was itself a Go rewrite of ownCloud.
I could use something more lightweight, focused, and easy to host on Kubernetes than Nextcloud.
How easy it is to host? How reliable is it? How good are the mobile apps?
Returning to the Valley from SF, feeling mostly quite tired (in a generally good way). Will soon depart, then work, then sometime in the evening decided whether to install a MikroTik CCR2004-1G-2XS-PCIe, or a Mellanox CX-5 2x25G (OCP v2 form factor) mounted on a PCIe converter card. 🤔
Ok, suppose it's time to get ready.
#homelab #SF #networking #linux #freebsd #engineering #nvidia #mellanox #mikrotik
Our public mastodon instance is up and running!
Exposed all these through a Pangolin Tunnel on a $5 Linode VPS in Singapore. Using S3 for offsite backups. Next up, a Mastodon instance at social.buddyverse.net! Not sure what to do with the main domain yet, maybe just a page linking all services.
I finally got buddyverse.net rolling from my homelab. Dropped $650 on some sweet refurbished HP EliteDesk and Dell OptiPlex mini PCs. Got two Proxmox servers up, planning a third for OPNSense firewall soon. Here's what's live:
1. Lemmy: https://lemmy.buddyverse.net
2. SearXNG: https://search.buddyverse.net
3. Pastebin: https://paste.buddyverse.net
4. Redlib: https://redlib.buddyverse.net
5. ConvertX: https://convert.buddyverse.net
6. Rustpad: https://pad.buddyverse.net
boostedOk, I still need a task app to replace google keep Problem is I have yet to find anything that meets these requirements * Able to be self hosted I have yet to find something that meets all of these requirements, maybe there has been some polish in the year since I looked last?
* Native android app
* Fast/smooth
* must work when offline, and sync seamlessly
* Must support reminders, including recurring, location based reminders a big plus
* Must have a very minimal interface, with the option to add checkboxes/lists
I turned a tiny #Mac mini into my own #Mastodon server—secure, low‑maintenance, and running like a dream. Here’s the full build, from #Docker to #Cloudflare Tunnel, and why #SelfHosting doesn’t have to be a headache.
Has anyone experienced this? Could this be a networking issue, a storage bottleneck, or something else? I’d appreciate any insights.
(5/5)
#sysadmin #system #homelab #selfhosting #linux #proxmox #opensource #tech
I had to force shutdown the server and restart. I get why the PBS VM might crash - 4 GB RAM, 2 CPU cores, and my internet speed maxes out at 60-70 Mbps - but why does the host server itself hang?
(4/5)
#sysadmin #system #homelab #selfhosting #linux #proxmox #opensource #tech
Here’s the problem: when I try backing up a VM from another server (Proxmox 1), the VM is 300 GB but only ~30 GB used. Everything is fine until it hits 34% (104 GB), then the PBS VM and the host server become unreachable.
(3/5)
#sysadmin #system #homelab #selfhosting #linux #proxmox #opensource #tech
I run Proxmox Backup Server (PBS) in a VM on one of my servers - let’s call it Proxmox 2. PBS uses 100 GB of cache storage because I’m using S3 as a datastore. Chunks are stored in the cache before uploading to S3.
(2/5)
#sysadmin #system #homelab #selfhosting #linux #proxmox #opensource #tech
I solved one issue, and now two more popped up. My Proxmox backups were super slow because of my router. I got a Netgear GS108E Gigabit switch, and now backups are way faster. 🎉
(Please boost 🙏 .. Need urgent help)
(1/5)
#sysadmin #system #homelab #selfhosting #linux #proxmox #opensource #tech
Fourth, I want to enable hardware acceleration on the EliteDesk for a VM running Jellyfin (in Docker) using the i7-9700’s UHD 630 iGPU. The Proxmox docs are vague on Intel GPUs, so I need a clear guide specific to this CPU. The networking issue is the most frustrating - has anyone seen a bridge drop connectivity when a VM stops? Any suggestions for troubleshooting or fixing this? I’d really appreciate pointers to resolve these problems.
I’m new to Proxmox and struggling with my homelab setup on two machines running Proxmox 9: an HP EliteDesk 800 G5 Mini (i7-9700) and a Dell OptiPlex 7070 Micro (i3 9th gen). First issue: on the EliteDesk, I have two Ubuntu Server 24.04 LTS VMs sharing a bridge. If I stop or shut down one VM, the other loses internet access, though I can access applications locally. Any ideas why the bridge fails like this? What can I check to fix it?
1/4
Buddyverse Status Page: https://status.buddyverse.net
Expect some downtime today while I'm applying thermal paste.
THIS is Why I Bought a #Framework Laptop | Hardware Haven
https://www.youtube.com/watch?v=mAYKlKyv6tI
tl;dw: upgrade #Framework13 laptop, slap old motherboard - which fully supports this - in a 3D printed case, and do any number of things with it
entertainment PC, mini desktop, blade server in a cluster, NAS with 6x SATA HDD ... take your pick!
maybe that high up-front cost isn't quite so bad after all?
🙅 Goodbye Forever OPNsense 🙅
It displeases me to finally and heartily say GTFO to OPNsense; to abandon a solid decade of use.
I've had it on everything from embedded arm64 experiments to baremetal with ranges of 10, 25, 40, and 100GbE NICs. I've used all of the core features, built complex global service meshes, H/A systems, etc. I used to love it. I used to pay for it.
OPNsense was great, until it wasn't (starting around the time they axed their use of HardenedBSD), and with each release it gets more convoluted, out of date, tedious to debug, and generally a source of disappointment. The command line controls are anemic, inconsistent, and the lack of unified and useful system state tracking is a source of sailor level obscenities. Also, dear gods get rid of XML configs, no one can parse it without going blind! What is this, SOAP and XML-RPC era nonsense, really? 😠
I do not have time to waste, and I do not say that lightly.
I am never debugging OPNsense ever again, especially not for four hours on a (yesterday) Saturday, and especially not putting off updates in a colo for TWO YEARS because their team decided to break admin group SSH controls, hamper CARP flapping controls, breaking IPMI fencing, and the list goes on. I am done.
What now? Three realistic options.
1) BSD Router Project: I've built custom BSD-RP releases with Poudriere, loved just about everything it offers.
2) VyOS: configurable via CLI in a fraction of the time that was wasted on debugging OPNsense. Solid product, enjoying it more every day.
3) OpenWRT: I build custom releases for NanoPi and Meraki rooted WAPs and SOHO boxes, it's fun, though it's not running my 100G infra.
#opnsense #bsd #freebsd #linux #networking #engineering #homelab
Am going to try out #Exosphere this evening on my four #homelab machines - looks super interesting, especially since I’m so bad at keeping machines up to date. 🙈
See https://untrusted.website/@mr_daemon for related Toots.
Was jokingly nerd sniping a friend, and then failed my own WIL check and put together my own mini rack. Bit top heavy so either need some temporary weight in the bottom or to move things around a bit, but kinda fun mini project to put together. (and cleans up a rats nest of cables on my desk)
Also have some shorter usbc cables coming tomorrow which should let me tidy up a bit more.
🌐 HomeLab DCIM Rezoning 🌐
Circumstances lacking an interesting story, my Netbox instance (via FreeBSD jail) has been offline during the year of relocations from SFO to Vancouver to Chicago to Sunnyvale. Fun!?
Anyway, I'm resyncing the regions, subnets, sites, racks, and generally resuming a substantial amount of infra/dcim maintenance over the past several days. Housekeeping, basically.
Perhaps at this point the term "HomeLab" is a stretch. There are eight sites across two continents, six colos, five racks I can touch in person, a few hundred switch-ports up to 1-100G, a bit over 1PB of storage, and a MRC equivalent to a 4bd house mortgage in most of the USA. 🧐😲
Perhaps it's time to start a 503.C and share these resources more with the BSD community. Probably so. ❤️
Fun #homelab stuff today. You know how Linux numbers drives /dev/sda, sdb, etc? For the first time I’ve seen /dev/sdq. It’s one of these Dell’s with 26 disk slots. I only have 15 in there at the moment. It also has 288G of RAM. It’s my biggest system so far.
“Hey Griswald? Where ya gonna put a server that big?”
“Bend over and I’ll show you.”
I’m actually organised as a tax exempt, non-profit micro ISP in the US. Just getting it off the ground. Going to offer internet services to individuals and other tax exempt orgs. This will be one of the servers.
#selfhosting
Furries and Gentlegeeks, stop what you’re doing. Reddit pointed out the holly reason for 3D printers to exist: The 6-bays ThinkNAS
Behold my latest #HomeLab MacGyver, inside my Brocade ICX 7250 PoE, a fan presently held onto a heat sink with a pair of elastic bands. (Yes it's a temporary fix, but we all know how permanent they can be, especially when getting the elastic bands round the heat sink without scraping any of the surface components around it was such a git)
Yep. Spinning rust drives do take up a bit of power. I have 12 x 2.5" SAS drives and they're all formatting (used NetApp drives are cheap and plentiful on eBay, but they come formatted to 520-byte sectors. Takes like 4-5 hours to reformat to 512-byte sectors).
The spike up is when the system went from 6 to 10 drives. (the last 2 were fiddly, so they came in a couple minutes later)
Hmmm, ok, today's #HomeLab challenge for myself, find out what the hell I've fucked up to cause this level of asymmetry:
iperf3 -P10 --bidir -c 10.254.254.109
[...]
[SUM][TX-C] 0.00-10.00 sec 262 MBytes 220 Mbits/sec receiver
[...]
[SUM][RX-C] 0.00-10.00 sec 2.87 GBytes 2.47 Gbits/sec receiver
10x faster in one direction than the other O_o
Je parle un peu de mon infra auto-hébergée en ce moment et des changements que je fais dessus. Voilà ce que ça donne en photo pour la partie principale.
À ce stade, ça ne devrait plus trop bouger.
OK. Step 1 done. My homelab machines get ULAs and I have added AAAA records to my DNS. So this already works :) Ping and ssh via IPv6 with hostname resolution via my own DNS server. Now to add reverse lookup.
Unfinished brain dump at https://codeberg.org/jwildeboer/gists/src/branch/main/2025/20250827BasicIPv6Unfinished.md
Corrections and enhancements welcome! I'm a noob wrt IPv6 ...
3/4
#ItsAlwaysDNS Finally learning IPv6, a thing I tried to avoid for the longest time. All my servers are reachable via IPv6, I know how to do that, but now I am learning to set up IPv6 in my home(lab) network and wrapping my head around concepts like ULAs (Unique Local Addresses, that don't route to/from the internet), prefixes I get from my ISP and how all of that translates to DNS entries on my BIND9. Mistakes will be made.
1/4
I was looking at how to publish docker images generated by forgejo runners to docker hub (or something else) then I found this blog post https://mteixeira.wordpress.com/2025/02/03/my-self-hosted-forgejo-runner-setup/#creating-a-registry-push-secret
At the end, I saw a link to a Mastodon account and said, hey, this blog post is awesome, let's follow this person.
Guess what, the author is @badnetmask and I already follow him 😂
Nice job!
Getting 100MB/s when transferring a huge directory of tiny files (nextcloud photos/thumbnails) to the Ceph cluster via rsync daemon, not bad tbh
#Kubernetes #Ceph #Homelab
At home my Fritz Box takes care of DHCP and stuff, so laptops, phones etc. are in the fritz.box default domain. My homelab however has its own domain, homelab.jhw. So I have a mixture of hostnames with different domains in the same subnet. I just upgraded my home server, srv01.fritz.box and with this DNS setup, I could get a certificate with certbot for my home server from my own CA, which runs as a container on the Raspberry Pi:)
2/3
#ItsAlwaysDNS I took the plunge. Configured my Fritz Box to hand out my new DNS server to all DHCP devices. So now DNS requests go first to my BIND9, which resolves homelab.jhw. If it can't answer, it forwards to the Fritz Box, who takes care of the local fritz.box domain. And if that doesn't work it goes out to the internet. Seems to work :) I can now reach my homelab machines under their homelab.jhw name and with a valid certificate. Yay! Updated the gist too.
3/3
Been printing #gridfinity bins for IKEA Kallax to store my electronics. Sorting random components now. Wondering if I made a dumb call: merged capacitors by value instead of brand/voltage/value. Easier storage, but bad idea? I’ll check specs when using. #electronics #makers #3Dprinting #homelab
#ItsAlwaysDNS This is rather advanced stuff, but I brain dumped how I configured my Bind DNS server for my homelab that allows all my servers, containers etc to resolve local and global names. Just the config, no detailed explanations. Might do that as a blog post. Should I?
https://codeberg.org/jwildeboer/gists/src/branch/main/2025/20250826DNSHomelabBind9.md
UPDATE: Blog post done, see https://social.wildeboer.net/@jwildeboer/115113695729233240
1/3
| Yes, explain Bind!: | 104 |
| No, it will go wrong: | 1 |
| Just the results: | 4 |
Closed
Hardware Haven pokes at one of those fanless Intel N150 systems with 4x 2.5GbE
it works with 32GB memory (16GB officially supported), has an NVMe slot, and other interesting features such as an RJ45 serial port and some provisions made for 4G or 5G modems.
Anybody running a #Dell #idrac server in their #homelab? I bought one used recently and it came with a system name set to WIN-AFGAQLM38BR. I want to change the hostname. But all I have is the idrac web interface (or ssh to it) and linux OS (XCP-ng). How does one set that hostname?
Also the service tag has been set to 1234567. That surprises me. I didn't realize that could be set.
@NekoArc On bare metal you have 2 great choices. Xcp-ng, managed by xen orchestra, and proxmox. I use xcp-ng and lots of folks use proxmox. Once you do that, all the real work is done in VMs. And maybe you have one VM that runs something like Debian and then runs containers in it. Set up a different VM for logging in and doing your your every day, interactive mucking about. Virtually never log in on the bare metal OS. Leave it be.
I owned one of those HPs for about 15 minutes. It was always loud. Never got quiet. My Dells slow the fans to a whisper unless they get hot.
(I keep them in the garage. They get hot. 😜)
#homelab #selfhosting
And, following my own guide [1], I now have step-ca as Certificate Authority up and running as container (pod) on the Raspberry 4 for my homelab.
[1] https://jan.wildeboer.net/2025/07/letsencrypt-homelab-stepca/
#ItsAlwaysDNS but I managed to get my own BIND9 name server up and running on the Raspberry 4 in my homelab. So I can now continue with setting up a new Certificate Authority (CA) as container. (please refrain from replying with "use dnsmasq!" or other solutions. I know what I am doing and I am stubborn. I'm a BIND9 guy. Period. :)
TIL (Today I learned): If you wonder why your PoE (Power over Ethernet) Splitter fails to power the Raspberry 4 in your homelab, it helps to check the cables again. TL;DR my PoE injecting network switch has 8 network ports that deliver PoE but also two uplink ports that don't. Guess where I put the network cable ;) Solved. And now my Raspberry Pi 4 with Fedora 42 is connected with 1 Gbit/s, PoE powered and can be configured as DNS server :)
Just had my first power outage since getting the Beelink mini nas and found it hadn't turned back on when power came back.
I had looked in the bios when I set it up, but didn't find anything, but the last post in this thread explains which option to change.
https://bbs.bee-link.com/d/7075-me-mini-does-not-auto-power-on-after-power-failure/5
Proxmox in Enterprises: I'm often asked, 'Can we use our Active Directory, LDAP, or OIDC with Proxmox?' Yes, you can!
Let's have quick dive into installing and configuring Authentik and configure Proxmox VE to use OIDC as an additional authentication realm.
#Proxmox #ProxmoxVE #opensource #Authentik #OpenID #OpenIDConnect #OIDC #EntraID #enterprise #homelab
https://gyptazy.com/proxmox-authentik-oidc-install-configure-and-connect-authentik-to-proxmox-ve/