For those unfamiliar, Open WebUI is a self-hosted AI interface, which you can use with local models with Ollama, OpenRouter, etc.

Also note there was a recent license change, which is why I didn't say 'Open source'. You can make your own judgements about that here:

https://docs.openwebui.com/license/

See the link in the post for all the changes, there were too many to list and lots of quality of life improvements from what I can tell.

I learned about Appwrite from the entry in the selfh.st newsletter. It looks like an open source (BSD-3 Clause license) alternative to Google's Firebase and other BaaS platforms:

https://appwrite.io/blog/post/open-source-firebase-alternative

I haven't tried it, but it looks neat

Hi,

As the title suggests: what are alternatives to syncthing that are basically fire and forget, works on multiple device types, and just focuses on file syncing?

I've had over the months the weirdest problems with syncthing, and lately I noticed some of my photos got corrupted, which is an absolute no no for me. I use syncthing currently as a easy automatic backup of documents, photos and other files, between my PCs and my phones (they all send only to the server. Folders are not shared with other devices).

So currently I am running a 4 bay Synology DS423+. I’ve upgraded it to have 32gb ram and a 10g network port. I am looking to upgrade to something with more bays like 12+. I was looking at the DS2422+ since it is 12 bays and I could transfer my ram from my current synology to it.

Now if I build one myself something I’ve never done is it easy to just move my storage over and not lose any data and be able to access it? I am running 4x 12TB in raid 5 in my current. It’s mainly my backup, plex files and also running qbit and tautulli. I saw some people recommend 45homelab HL15 would that be a good swap?

I'm looking for a compact NAS to back up events from my video surveillance system. Two drive bays is enough, maybe four at most. They can be 2.5" or 3.5", SATA or SAS, preferably populated with mechanical drives but even with reliable SSDs. It doesn't need to handle more than a few GB per day of throughput and 16TB of total storage would be more than enough so it doesn't need to support even more massive drives. I don't care if it's complete product like a Synology or something built from scratch using an SBC and adapters; all I need is RAID 1 and an SMB/CIFS file share, though I would like to keep costs low. My house is wired for Ethernet so wifi would just be a bonus but it might help to hide the device somewhere a burglar isn't likely to see it like they will the NVR in my server rack. Also, a GNU/Linux-based OS is obviously mandatory or else I wouldn't be on Lemmy.

There are probably some of us still sitting on free ESXi hosts, but as I've gotten off mine to Proxmox ive been pretty pleased so far. Leave ESXi, it was a great way to learn but Proxmox is more better.

Here is the one big trick I learned: if using the native import tool, shutdown the source VM and delete all the snapshots! Each snapshot adds an exponential slowdown due to the crappy API. If importing windows, remove the VMWare Guest Tools first. That seems to make the import process work pretty well. Linux and FreeBSD don't seem to care but Windows sure does.

Good day all. Looking for a self hosted Trello for BT Free. I've found:

https://github.com/plankanban/planka

and

https://github.com/michelematteini/tarallo

Does anyone have experience with these or does anyone recommend something else that they have personal experience with?

Thank you!

Curious about thoughts on Garage as an alternative to Minio. It has been in development since 2020. Here is the project git. Documentation looks nice.

Curious what others think of it as a project that has been around for a few years and seems like a solid, open source contender now that Minio has removed most of their community edition functionality.

Hi everyone!

I recently created a user-friendly WiFi portal that runs on my private LAN. The goal was to provide my family with quick access to WiFi information without them needing to ask me for it. After searching for a similar solution and not finding one, I decided to build it myself!

Some of the key features:

• Login info hidden unless you explicitly click on it.
• Generates QR code for your use.
• Mobile and PWA support.
• Tags to help identify when to use each network.

I'm not a web developer, so I would really appreciate any feedback or suggestions on how to improve the app. Thanks for your support!

https://codeberg.org/ch8zer/wifi-portal.git

Hi everyone,

I’m a PhD student in Computer Science researching why people choose to self-host software—what motivates you, what concerns you, and what factors affect your decision-making.

To better understand this, I’ve prepared a short anonymous survey (~10 minutes). Your insights as part of the self-hosting community would be incredibly valuable for this research.

🔗 Survey link: https://survey.lpt.feri.um.si/376953?newtest=Y&lang=en&s=ls

This study is part of my doctoral research at the University of Maribor, Slovenia, conducted under the supervision of Assist. Prof. Lili Nemec Zlatolas, PhD. All responses are anonymous and used strictly for academic purposes.

If you’ve ever self-hosted anything—or even just considered it—I’d really appreciate your input.

Thanks a lot for your time, and feel free to ask me anything about the project ([email protected])!

Cheers!

Given that Watchtower is potentially unmaintained now, this might be a cool alternative?

Screenshot:

Features from their github:

• Extremely fast. Cup takes full advantage of your CPU and is hightly optimized, resulting in lightning fast speed. On my Raspberry Pi 5, it took 3.7 seconds for 58 images!
• Supports most registries, including Docker Hub, ghcr.io, Quay, lscr.io and even Gitea (or derivatives)
• Doesn't exhaust any rate limits. This is the original reason I created Cup. I feel that this feature is especially relevant now with Docker Hub reducing its pull limits for unauthenticated users.
• Beautiful CLI and web interface for checking on your containers any time.
• The binary is tiny! At the time of writing it's just 5.4 MB. No more pulling 100+ MB docker images for a such a simple program.
• JSON output for both the CLI and web interface so you can connect Cup to integrations. It's easy to parse and makes webhooks and pretty dashboards simple to set up!

We’re excited to announce the release of Stalwart v0.12, a significant milestone that evolves Stalwart from a powerful mail server into a complete, integrated communication and collaboration platform. This release delivers one of the most anticipated features from our community: native support for calendars, contacts, and file storage—all built directly into the server, with no need for third-party integrations.

I've been doing more with voice commands on computer, and wanted to find an alternative to Microsoft voice assist. Talon goes way deeper than that though, you can use it as as ample dictation, but some more advanced users can basically do almost everything with just their voice and do stuff including launching apps , built on python, and it has it ton of amazing functionality. And especially great blessing for people that are older have problems with their hands, and combines nicely with a few other open source apache licence software

This one is less focused on self-hosting a homelab service, but I thought might be interesting for the homelabbers here. I got into this hobby through my career in cybersecurity, and decided to write up a little post about a tool I frequently use, mitmproxy!

I run my production Jellyfin server and a few other services on a Optiplex sff computer with a thicc hard drive and a low profile GPU.

I want to build two more of these with thicc Hard drives so that my parents and my in-laws can have a local Jellyfin instance that I manage remotely and they just need a box plugged in somewhere at their homes.

Is it possible to make Proxmox build a VPN tunnel on boot so I can just have it in my cluster dash. Like using tailscale or openvpn.

Or am I going to have to go with my original plan and put that on the same box as the Jellyfin server and then just VNC in?

Any tips or ideas?

I've started using beets to manage my music library, but it doesn't work well with jellyfin. As you can see, it creates about a million artists off of features, and this makes it hard to use. I can't find a way to fix this in beets, so I'm considering switching, but haven't found any proper alternatives. Do you guys solve this in any way, or use a different management software that is more standard? Thank you!

Hi,

I've been been happily running paperless for a while now, and I've even got a few useful mail actions. However, I'm having trouble keeping the oauth2 (gmail) configuration alive past a few weeks. I suspect the refresh token expires, since the configuration stops working. And since I can't figure out how to re-authenticate, I have to create a new configuration, move all actions to the new one, and delete the old one.

Does anyone know how I can keep the original config and just re-authenticate?

Thanks

Hi,

I would like to host my own email server.

I know the pros and cons of doing so, please do not post about the advantage or disadvantage, there is others open topics for it.

I would like my setup fit those "requirements"

• FLOSS
• Be able to create complex rules for incoming/outgoing email ( with 🐍 Python language preferably ).
  • based on any properties (subject, ip, header, size, attachment, signed, encrypted, etc.. )
  • able to modify, delete, fw, etc... the email
• Manage the limit of sending by size and number of outgoing over time (min, hours, week, etc..)

If you know a matrix comparison between email servers I'm all ears.

If you have any feedback with a solution this is most welcome too.

I plan to update this post to create a matrix/spreadsheet of all the solutions that I'll gather.

Thanks.

Hi,

It had been twenty year that I stopped my couples of self-hosted email servers.. ( That did run on a 10 years span )

Now, I'm digging to relaunch one.. OMG the GAFAM etc... did well screw us !!

I think those two post summarize well what happened...

On the technical level email are OLD ! ~1982(SMTP), and since then few revisions were released, but they only build ~~extra thing~~ complexity on top of it !! and the last revision date was in 2008 ! ( 17 year ago... )

And they are complex because of this build-up,
For the example, the list of the daemons running in docker-mailserver give a clue...

• Postfix
• Dovecot
• Rspamd
• Amavis
• SpamAssassin
• ClamAV
• OpenDKIM
• OpenDMARC
• Fail2ban
• Fetchmail
• Getmail6
• Postscreen
• Postgrey
• Support for LetsEncrypt, manual and self-signed certificates
• SASLauthd with LDAP authentication
• OAuth2 authentication

On the mass level, the GAFAM managed to convince the mass that email server (and more broadly any self-hosted (aka computing) ) is complicated, so "let's us do" that could be understand as "Let's us own your technology"

For a time I was thinking "maybe I should get away from email, that only belong the GAFAM now... and maybe found an alternative... ?" But If I found an alternative, I must convince the others to do the same... slower... way slower...

No ! , the first step is to have more and more people re-owing their technology ! So having more and more self-hosted email server again..

To reverse the tendency, instead of feeling like a black sheep (and be censored) to not have a GAFAM email. It will be people that use a GAFAM email that will pointed out ! to have deleted ( or move email to SPAM without reason etc..) your email from [email protected]

If you use a none GAFAM email ( like me ), and someone tell you:
"hoo sorry I didn't get it"
"Sorry, I didn't see it, it felt in my SPAM folder" (with a tone that's your fault because you use something else than everyone else (aka GAFAM))

Please note, that legally, is their responsibility ! Whenever it was automated or not !
If your MTA[^MTA] did send your email the the recipient MTA it's their sole responsibility...\

and if the attempt has been blocked before reaching the destination MTA, by a firewall or something else on their side (even on ISP level), no matter if they own it or not, it's also their responsibility :) )

[^MTA]: Mail Transfer Agent Handles the transfer of emails between servers using SMTP

Hi everyone, I am trying to repurpose a Ryzen 1700 system for a home server, but not exactly sure what the best solution for my needs is, and how to find additional resources.

More context, I have 4 8th hdds (wd blue drives; would’ve preferred reds but, alas). I intended to run these in raid10, but open to other ideas also. These are connected via sata directly to Mobo. I’d like to selfhost a nice NAS stack, to include: my own office 365 / google docs thing, file storage, and storage and playback of music and video files. I’d like to run jellyfin and a myriad of ‘arr things. Please send any and all suggestions. Should all of these run on a single virtual machine?

Alongside this, probably in a separate virtual machine, I’d like to run a home assistant instance with some mild transcoding (I think) going on in regards to some cameras I have around the house.

I also think I’d run tail scale to vpn back in?

What I’ve researched so far is proxmox and casaos (lightly). Casaos is alluring mostly because it seems like an easy on ramp, with lots of visual configuration. I enjoy CLI config, but visual configs are easier to discover settings and options that might not occur to me. I’d ideally favor stability here, as I like to tinker, but don’t have a huge amount of time for it.

Am I on the right track with all this? Any pitfalls? Any must have self hosted software I should be sure to include? Should I set up the storage pool in proxmox first as raid10? Any general advice? Words of encouragement? I’ll take it all.

Apologies if this is the wrong forum — if so, please feel free to delete (and direct me hopefully to a more appropriate locale).

Edit: forgot to mention, system also has a slower ssd boot drive, and a 1070 I plan to pass thru as needed.

I've had a few of the *arrs running in Docker on Unraid without issue for several months now. Yesterday afternoon I was ripping a bunch of CDs that my wife had picked up, and midway through I noticed that Lidarr was not allowing to add new artists or albums that were not already populated in my library. Manually searching for anything results in the attached screencap and the log entries below:

[Warn] HttpClient: HTTP Error - Res: HTTP/2.0 [GET] https://api.lidarr.audio/api/v0.4/search?type=all&query=billie+holiday: 500.InternalServerError (33 bytes) {"error":"Internal server error"}

[Warn] LidarrErrorPipeline: NzbDrone.Core.MetadataSource.SkyHook.SkyHookException: Search for 'billie holiday' failed. Unable to communicate with LidarrAPI. [v2.11.2.4629] NzbDrone.Core.MetadataSource.SkyHook.SkyHookException: Search for 'billie holiday' failed. Unable to communicate with LidarrAPI.

I've done a little searching, and I gather this is a result of my Lidarr instance not being able to communicate with a metadata server, but the troubleshooting advice seems to be focused on new installs with network configuration problems. I had made no configuration changes during the ~20 minutes since my last successful search (nor really in the last couple weeks), and all my other *arr apps are working perfectly fine, so I'm kind of at a loss. I've tried restarting and updating the container, but no dice, so here I am...

Ideas/suggestions?

I currently have a Synology 220+ and a couple of VPS's, and I'm looking to consolidate, while getting out of Synology's walled garden. I've already got a couple of 3.5's in the Synology, and 4 2.5's lying around and I'm planning on running a number of docker containers and a couple of vms.

That said, I've never built anything before, and basically just went to PCPartPicker, started with the case, and checked 5-stars on each component and went from there. So... how absurd is my build?

PCPartPicker Part List

Partially Solved

While I haven't found a native solution on how to integrate NTFY to glance, I did build up something that actually send basic text streams to glance in an automated way. It's very rudimentary and probably error prone, but that's the best I could do right now... Maybe someone else will chime in and give some better advice/solution.

For those interested postgREST allows to build a simple docker container postgres database you can query for the custom api in glance. It DOES work, but If like myself, your database/json/postgre knowlege is very limited, it only allows basic text response like: "Update Failed".

I did try to get a little further into the rabbit hole, but it does come with the necessity to have a good database and query/response background ? Not a very good solution and will probably not go one or try to improve on that right now... But feel free to give better advice or another lead to follow :)

Further notes:

• https://docs.postgrest.org/en/v12/tutorials/tut0.html
• https://samkhawase.com/blog/postgrest/
• https://deepwiki.com/glanceapp/glance/6.1-template-functions
• https://github.com/glanceapp/glance/blob/main/docs/configuration.md#custom-api

On a final note, I do see a lot of interest in the Glance community and alot of new and interesting updates:

• Added .Options.JSON to the custom API widget which takes any nested option value and turns it into a JSON string v0.8.3
• [Custom API] Synchronous API calls and options property v0.8.0

Hello everyone !

I kinda hit a roadblock here and I'm interested if someone actually have done something similar or an alternative to what I'm trying to achieve.

Some background

Right now I'm playing around with NTFY and works great. I even hooked some automated backup script to my server with stdout/stderr output:

(Please, no bash-shaming ! :P)

#!/bin/bash

$COMMAND

if [ $? -eq 0 ]; then
        echo "Success"
        issue=$(<stdout.txt)
        curl -H "Title: Hello world!" -H "Priority: urgent" -d "$issue" https://mydomain/glancy

else
        echo "Failure"
        issue=$(<stderr.txt)
        curl -H "Title: Hello world!" -H "Priority: urgent" -d "$issue" https://mydomain/glancy

fi

This works great and I receive my notification on every device subscribed to the topic

What I'm trying to achieve?

Send the NTFY notification to a visual dashboard like Glance. If there's no native way to achieve this, self-host a simple json api that get's populated by my server's script response?

What's the issue ?

After skimming all the GitHub repos, there's no mention on any self-hosted dashboard to integrate NTFY as a notification hook. I find it kinda strange because NTFY is just a simple HTTP PUT or POST requests so It should be rather easy no?

And after searching the whole day on the web, there wasn't any good results or resources. So I came to the conclusion that It wasn't that easy and probably needs a bit more of something I'm probably bad at (coding?).

In the glance documentation there's configuration to hook a custom api and looks rather simple, however now I hit a roadblock I'm not able to solve... I have no idea where or how to spin up a self-hosted and dynamic json api that communicates with my server and updates/populate that json file... Here's an example to show what I mean:

Json api: https://api.laut.fm/station/psytrancelicious/last_songs

Custom Glance API template:

- type: custom-api
  title: Random Fact
  cache: 6h
  url: https://api.laut.fm/station/psytrancelicious/last_songs
  template: |
    <p class="size-h4 color-paragraph">{{ .JSON.String "title" }}</p>

Questions

1. Any native way to hook NTFY's notification to a dashboard like instance (Glance, Homer, Dashy?)

2. If no, Is it possible to self-host a json api that gets populated by my script's response? A good pointer to the right direction would be very nice, preferably a Docker solution !

3. Another solution to have a visual dashboard (not the native NTFY dashboard) and visualize all my script response notification in one place ?

Thank in advance for all your responses :) and sorry for my bad wording, web development terminology is not really my cup of tea !