21

[AIT] what is your selfhosted local LLM stack?

preview

What tools and projects do you use for your local LLM chats and agents?

I’m trying to sift through the mess of LLM harness and tools to find how to really get the most out of local LLMs.

What tools and projects do you use for your local LLM chats and agents?

I’m trying to sift through the mess of LLM harness and tools to find how to really get the most out of local LLMs.

84

Anybody running IncusOS here?

Linux Containers - IncusOS - Introduction linuxcontainers.org
preview

I discovered Incus and IncusOS recently, and they seem like a very cool foundation for a self-hosted setup.

I’m thinking of using them with incus-compose to declare my setup.

Does anybody use them? I’d love to get some feedback. Are they stable and polished yet, or are there still a lot of sharp edges?

They mention “designed for modern systems” on their website. Are there any issues running it on a roughly 10-year-old laptop?

I discovered Incus and IncusOS recently, and they seem like a very cool foundation for a self-hosted setup.

I’m thinking of using them with incus-compose to declare my setup.

Does anybody use them? I’d love to get some feedback. Are they stable and polished yet, or are there still a lot of sharp edges?

They mention “designed for modern systems” on their website. Are there any issues running it on a roughly 10-year-old laptop?

88

There isn't an easy exploit for this PHP vulnerability

Elias Taylor There isn't an easy exploit for this PHP vulnerability eliastaylor.com
preview

I came across the link via https://mastodon.social/@CuratedHackerNews/117481600704689796

Last time I touched php it was ~20 years ago and maybe for a month. Could someone with actual php understanding weigh in?

Somehow, somewhere between the lines, my understanding was that there have been observances of the vuln being used in the wild before the 14 UTC. And that wasn’t from reading news sites

I came across the link via https://mastodon.social/@CuratedHackerNews/117481600704689796

Last time I touched php it was ~20 years ago and maybe for a month. Could someone with actual php understanding weigh in?

Somehow, somewhere between the lines, my understanding was that there have been observances of the vuln being used in the wild before the 14 UTC. And that wasn’t from reading news sites

25

How do you all handle port 80 and HTTP-01/DNS-01 challenges for Let's Encrypt cert renewals?

preview

I’m currently running a bunch of services behind a reverse proxy and that proxy is configured to use let’s encrypt for ssl certs using DNS-01 challenges for renewals. All of the services are configured to only work using Tailscale IPs, so, as far as I’m aware, none of the services are open to the internet, except for port 80.

I noticed if I don’t leave port 80 open, when it’s time for renewal, everything fails. This leads to a bunch of logs on my Unifi setup saying intrusion attempts have been blocked on port 80. I thought I could go to Let’s Encrypt’s website and find a list of IPs I could put on an allow list but they don’t post those. So my questions are:

1. If I understand the process correctly, Let’s Encrypt checks the DNS TXT record to verify I own a domain, then sends a new cert to my server over port 80; meaning, I have to have port 80 open during the renewal process, correct? Nope

2. If that’s correct, what do you all do about port 80? Do you just have a script to open the port during the renewal process and close it afterwards or do you just leave it open all the time? Is there maybe a tool that already has this functionality? not if using DNS-01 challenges

3. I think I have port 443 open as well but I don’t remember if that’s needed or if I just did that when I opened port 80 trying to get all my services working again. Does 443 need to be open too? not if using DNS-01 challenges

4. If there is a better way to do all this or if I fucked up somewhere, please share. This was just the first setup I got working. I fucked up, see below

Thanks

EDIT: After deleting the current certificate and going through the process of generating a new one, everything is working without port 80 open. I’m using NGINX Proxy Manager and I must have clicked on “Lets Encrypt via HTTP” and not “Lets Encrypt via DNS” when originally setting this up. Not sure how else this could have happened 😅

I’m currently running a bunch of services behind a reverse proxy and that proxy is configured to use let’s encrypt for ssl certs using DNS-01 challenges for renewals. All of the services are configured to only work using Tailscale IPs, so, as far as I’m aware, none of the services are open to the internet, except for port 80.

I noticed if I don’t leave port 80 open, when it’s time for renewal, everything fails. This leads to a bunch of logs on my Unifi setup saying intrusion attempts have been blocked on port 80. I thought I could go to Let’s Encrypt’s website and find a list of IPs I could put on an allow list but they don’t post those. So my questions are:

1. If I understand the process correctly, Let’s Encrypt checks the DNS TXT record to verify I own a domain, then sends a new cert to my server over port 80; meaning, I have to have port 80 open during the renewal process, correct? Nope

2. If that’s correct, what do you all do about port 80? Do you just have a script to open the port during the renewal process and close it afterwards or do you just leave it open all the time? Is there maybe a tool that already has this functionality? not if using DNS-01 challenges

3. I think I have port 443 open as well but I don’t remember if that’s needed or if I just did that when I opened port 80 trying to get all my services working again. Does 443 need to be open too? not if using DNS-01 challenges

4. If there is a better way to do all this or if I fucked up somewhere, please share. This was just the first setup I got working. I fucked up, see below

Thanks

EDIT: After deleting the current certificate and going through the process of generating a new one, everything is working without port 80 open. I’m using NGINX Proxy Manager and I must have clicked on “Lets Encrypt via HTTP” and not “Lets Encrypt via DNS” when originally setting this up. Not sure how else this could have happened 😅

19

Question(s) about self-hosting matrix server on Raspberry Pi 4 next to Navidrome and Jellyfin

preview

Hi everyone,

I am new to self-hosting and this is also my first post on Lemmy. I have a navidrome instance for my music (installed via docker compose) and a jellyfin instance for movies and shows (also installed via docker compose) running on a Raspberry Pi 4 8GB. I know I could also use music over jellyfin but I prefer to have it in navidrome.

The next thing that I want to achieve on my journey through the world of self-hosting is to run my own matrix server. The ideal eventual goal would be to have my own server on that same Raspberry Pi which I can then use to bridge Whatsapp to so I finally can get rid of that App from my primary phone. Receiving Whatsapp messages is unfortunately necessary for me to keep contact with certain people, although I would prefer to get rid of it completely as well, but who am I telling that, there are plenty of people out there facing the same issue.

As Whatsapp needs to run somewhere, which is required to have the bridge properly working, it would ideally run in an Android Virtual Device although that is no requirement as I know physical phones are recommended and emulated devices get blocked more often.

Ideally, the matrix server would also be installed using docker compose. It is intended to have one user only which is me but should of course be able to communicate with other servers to enable me to talk to others. The main limitation is that all three services should be running on that Raspberry Pi at the same time.

As my experience regarding such a project is virtually zero, my questions are:

Is such a setup is feasible and generally a good or bad idea (considering running jellyfin, navidrome and matrix on one Raspberry Pi)?

What issues should I consider before starting such a project?

If there’s someone having a similar setup, what are your experiences?

Also important to note: The matrix server should be able to handle video and voice calls as well, not in large rooms of course, but rather in a context of 2 to 4 people max.

Any advice, information or personal experience is well appreciated! I want to properly inform myself before starting out and thought here might be a good starting point. If I forgot to mention any specific information that is needed please let me know.

Thanks a lot!

EDIT: Adjusted post title to better reflect the advise-seeking character of this post and adjusted question phrasing

Hi everyone,

I am new to self-hosting and this is also my first post on Lemmy. I have a navidrome instance for my music (installed via docker compose) and a jellyfin instance for movies and shows (also installed via docker compose) running on a Raspberry Pi 4 8GB. I know I could also use music over jellyfin but I prefer to have it in navidrome.

The next thing that I want to achieve on my journey through the world of self-hosting is to run my own matrix server. The ideal eventual goal would be to have my own server on that same Raspberry Pi which I can then use to bridge Whatsapp to so I finally can get rid of that App from my primary phone. Receiving Whatsapp messages is unfortunately necessary for me to keep contact with certain people, although I would prefer to get rid of it completely as well, but who am I telling that, there are plenty of people out there facing the same issue.

As Whatsapp needs to run somewhere, which is required to have the bridge properly working, it would ideally run in an Android Virtual Device although that is no requirement as I know physical phones are recommended and emulated devices get blocked more often.

Ideally, the matrix server would also be installed using docker compose. It is intended to have one user only which is me but should of course be able to communicate with other servers to enable me to talk to others. The main limitation is that all three services should be running on that Raspberry Pi at the same time.

As my experience regarding such a project is virtually zero, my questions are:

Is such a setup is feasible and generally a good or bad idea (considering running jellyfin, navidrome and matrix on one Raspberry Pi)?

What issues should I consider before starting such a project?

If there’s someone having a similar setup, what are your experiences?

Also important to note: The matrix server should be able to handle video and voice calls as well, not in large rooms of course, but rather in a context of 2 to 4 people max.

Any advice, information or personal experience is well appreciated! I want to properly inform myself before starting out and thought here might be a good starting point. If I forgot to mention any specific information that is needed please let me know.

Thanks a lot!

EDIT: Adjusted post title to better reflect the advise-seeking character of this post and adjusted question phrasing

94

PHP selfhosting list

preview

Ive checked all my self hosted services and was happy to find out the only one was written in PHP, addy, formerly anonaddy.

You can use the awesome self hosted GitHub repo or this one already filtered for PHP

https://awesome-selfhosted.net/platforms/php.html

Ive checked all my self hosted services and was happy to find out the only one was written in PHP, addy, formerly anonaddy.

You can use the awesome self hosted GitHub repo or this one already filtered for PHP

https://awesome-selfhosted.net/platforms/php.html

50

Release FreshRSS 1.30.3 - patching severe PHP vulnerabilities

GitHub Release FreshRSS 1.30.3 · FreshRSS/FreshRSS github.com
preview

While you should probably take some services offline altogether, FreshRSS has an update out.

Just one of many updates that we’re going to be seeing over the next few hours / days.

For those who haven’t seen it yet

therecord.media/php-script-vulnerabilities-warnin…

Hundreds of websites taken offline as concern grows over two PHP vulnerabilities

huntress.com/…/php2shell-leaked-critical-php-vuln…

While you should probably take some services offline altogether, FreshRSS has an update out.

Just one of many updates that we’re going to be seeing over the next few hours / days.

For those who haven’t seen it yet

therecord.media/php-script-vulnerabilities-warnin…

Hundreds of websites taken offline as concern grows over two PHP vulnerabilities

huntress.com/…/php2shell-leaked-critical-php-vuln…

15

Mysterious Loop Devices in Disk manager on plex server

preview

Hey all, recently I was doing some work on the machine where I host my plex and audiobookshelf servers and I noticed that there were 3 new devices listed in the disk manager.

All 3 of the new devices have a description along the lines of “78MB Loop Device” and lists a backing file at location /var/lib/snapd/snaps/snapd_22291.snap (not exactly but they’re all similar to this). They show as unmounted and I’m not really sure what this is.

Additionally when I try to use libation to pull my latest audible books down it says it can’t read my local file location folder.

I recently saw the post about the plex vulnerability (which is part of why i was doing work on it, to make sure that my auto update was working) when I noticed this and it got me worried.

Are these new devices that showed up indicative that a malicious user has accessed my system through the internet and has setup a crypto miner to run in the background on my server or is it more likely that a disk is failing and this is the first sign?

Also before anyone feels the need to say it, I know that plex is not ideal and jellyfin is better, but I have remote users that I cannot talk through troubleshooting to get jellyfin when I can tell them to just install plex on their end.

Any help would be greatly appreciated

Hey all, recently I was doing some work on the machine where I host my plex and audiobookshelf servers and I noticed that there were 3 new devices listed in the disk manager.

All 3 of the new devices have a description along the lines of “78MB Loop Device” and lists a backing file at location /var/lib/snapd/snaps/snapd_22291.snap (not exactly but they’re all similar to this). They show as unmounted and I’m not really sure what this is.

Additionally when I try to use libation to pull my latest audible books down it says it can’t read my local file location folder.

I recently saw the post about the plex vulnerability (which is part of why i was doing work on it, to make sure that my auto update was working) when I noticed this and it got me worried.

Are these new devices that showed up indicative that a malicious user has accessed my system through the internet and has setup a crypto miner to run in the background on my server or is it more likely that a disk is failing and this is the first sign?

Also before anyone feels the need to say it, I know that plex is not ideal and jellyfin is better, but I have remote users that I cannot talk through troubleshooting to get jellyfin when I can tell them to just install plex on their end.

Any help would be greatly appreciated

93

Beginner looking for advice on open-source NAS hardware and OS for self-hosting

preview

Hey everyone! I’m completely new to self-hosting, but I’ve become convinced of its benefits and would like to get started in the near future. I’m still trying to figure out which hardware and software would be the best fit for my needs, so I’d really appreciate some advice.

What I’m looking for:

  • I live in the EU and want to move away from US-based and other non-EU cloud providers, which is one of the main reasons I’m interested in self-hosting.
  • I want to de-Google my digital life and replace services like Google Photos, Google Drive, and Google’s contact and calendar syncing with self-hosted alternatives.
  • I’d like to automatically back up photos from my phone, sync files, contacts, and calendars, and access everything remotely.
  • I want to avoid proprietary operating systems and vendor-specific apps as much as reasonably possible, ideally relying on open-source software.
  • (Eventually, I’d also like to consider hosting my own email server, although I’ve heard that this comes with quite a few challenges. This isn’t a priority for now.)
  • I don’t want to become a Linux or server administration expert just to keep everything running.

A little more context:

  • Users: Just two people.
  • Workload: Mainly file storage and syncing, photo backups and management, contacts, and calendars. No video editing, media transcoding, or other particularly demanding tasks.
  • Budget: Around €500–800 for the NAS itself, excluding storage drives.
  • Hardware preference: I’d prefer a compact, reasonably power-efficient, ready-made NAS rather than building a full PC, mostly due to limited space.
  • Technical experience: I’ve been building my own PCs for almost 30 years and consider myself an advanced Windows user. However, my last real Linux experience was with Ubuntu about 20 years ago.
  • Internet connection: 250/40 Mbps DSL with a public (dynamic) IPv4 address and a FRITZ!Box 7590 router, which supports WireGuard VPN.

Here are my main questions:

  1. Which operating system would you recommend? I’ve been looking into TrueNAS Community Edition and OpenMediaVault, but I’m open to other suggestions. Ease of use, stability, and low maintenance are important to me.
  2. Would a prebuilt NAS be a good choice? I’ve been considering devices like the UGREEN DXP4800 Plus and TerraMaster F4-424 Pro. Can their proprietary operating systems be replaced with something open-source without losing important functionality, such as fan control, power management, or hardware monitoring? Are there particular models known to work especially well with TrueNAS or OpenMediaVault?
  3. How much hardware performance do I actually need? Given my relatively modest requirements, would something like an Intel N100 already be sufficient, or would it make sense to invest in a more powerful CPU and additional RAM for future flexibility?
  4. Which software stack would you recommend? I’m currently considering Nextcloud for files, contacts, and calendars, and possibly Immich for photos. I’d prefer solutions with reliable automatic syncing and open-source Android apps.
  5. What should I consider regarding backups and remote access? I’m planning to use WireGuard for remote access, but I’d be interested in recommendations for a simple, reliable backup strategy as well.

I’m happy to learn new things and follow technical guides, but I’d prefer a system that doesn’t require constant tinkering once it’s set up.

Thanks for reading, and I appreciate any suggestions or experiences you’re willing to share!

Hey everyone! I’m completely new to self-hosting, but I’ve become convinced of its benefits and would like to get started in the near future. I’m still trying to figure out which hardware and software would be the best fit for my needs, so I’d really appreciate some advice.

What I’m looking for:

  • I live in the EU and want to move away from US-based and other non-EU cloud providers, which is one of the main reasons I’m interested in self-hosting.
  • I want to de-Google my digital life and replace services like Google Photos, Google Drive, and Google’s contact and calendar syncing with self-hosted alternatives.
  • I’d like to automatically back up photos from my phone, sync files, contacts, and calendars, and access everything remotely.
  • I want to avoid proprietary operating systems and vendor-specific apps as much as reasonably possible, ideally relying on open-source software.
  • (Eventually, I’d also like to consider hosting my own email server, although I’ve heard that this comes with quite a few challenges. This isn’t a priority for now.)
  • I don’t want to become a Linux or server administration expert just to keep everything running.

A little more context:

  • Users: Just two people.
  • Workload: Mainly file storage and syncing, photo backups and management, contacts, and calendars. No video editing, media transcoding, or other particularly demanding tasks.
  • Budget: Around €500–800 for the NAS itself, excluding storage drives.
  • Hardware preference: I’d prefer a compact, reasonably power-efficient, ready-made NAS rather than building a full PC, mostly due to limited space.
  • Technical experience: I’ve been building my own PCs for almost 30 years and consider myself an advanced Windows user. However, my last real Linux experience was with Ubuntu about 20 years ago.
  • Internet connection: 250/40 Mbps DSL with a public (dynamic) IPv4 address and a FRITZ!Box 7590 router, which supports WireGuard VPN.

Here are my main questions:

  1. Which operating system would you recommend? I’ve been looking into TrueNAS Community Edition and OpenMediaVault, but I’m open to other suggestions. Ease of use, stability, and low maintenance are important to me.
  2. Would a prebuilt NAS be a good choice? I’ve been considering devices like the UGREEN DXP4800 Plus and TerraMaster F4-424 Pro. Can their proprietary operating systems be replaced with something open-source without losing important functionality, such as fan control, power management, or hardware monitoring? Are there particular models known to work especially well with TrueNAS or OpenMediaVault?
  3. How much hardware performance do I actually need? Given my relatively modest requirements, would something like an Intel N100 already be sufficient, or would it make sense to invest in a more powerful CPU and additional RAM for future flexibility?
  4. Which software stack would you recommend? I’m currently considering Nextcloud for files, contacts, and calendars, and possibly Immich for photos. I’d prefer solutions with reliable automatic syncing and open-source Android apps.
  5. What should I consider regarding backups and remote access? I’m planning to use WireGuard for remote access, but I’d be interested in recommendations for a simple, reliable backup strategy as well.

I’m happy to learn new things and follow technical guides, but I’d prefer a system that doesn’t require constant tinkering once it’s set up.

Thanks for reading, and I appreciate any suggestions or experiences you’re willing to share!

300

PSA: PHP warns it plans to release 2 emergency patches imminently (today 14:00 UTC)

Ars Technica PHP warns it plans to release 2 emergency patches imminently arstechnica.com
preview

The set of vulnerabilities is dubbed PHP2Shell and disclosure is scheduled for 14:00 UTC with attacks expected to follow soon after.

It is adviced to take offline every non-vital services until the patch has been installed (or the Docker image has been updated with a patched version of PHP)

Update: fix is out www.php.net/archive/2026.php#2026-10-21-5

The set of vulnerabilities is dubbed PHP2Shell and disclosure is scheduled for 14:00 UTC with attacks expected to follow soon after.

It is adviced to take offline every non-vital services until the patch has been installed (or the Docker image has been updated with a patched version of PHP)

Update: fix is out www.php.net/archive/2026.php#2026-10-21-5

-30

[AIP] Was using a hermes agent to grab Internet Archive collections, had it automate the process and make a webui

GitHub GitHub - hardlysleeping/HardlyReading: Desktop web app for grabbing magazines, comics and PDFs from Internet Archive, with a library and reader. Komga-ready output. github.com
preview

Use at your own risk, wanted to take the ai out of the process so I wasn’t wasting hella tokens on repetitive actions. Search through and grab collections directly from internet archive. Still having the agent fix bugs and add features but it seems mostly stable now.

Useful since internet archive has a ton of old comics/magazines that you can’t find on comic repos like getcomics or batcave. Cracked, I couldn’t find anywhere else, tons of mad issues, carToons, every popular adult mens mag for the gooners, etc.

Use at your own risk, wanted to take the ai out of the process so I wasn’t wasting hella tokens on repetitive actions. Search through and grab collections directly from internet archive. Still having the agent fix bugs and add features but it seems mostly stable now.

Useful since internet archive has a ton of old comics/magazines that you can’t find on comic repos like getcomics or batcave. Cracked, I couldn’t find anywhere else, tons of mad issues, carToons, every popular adult mens mag for the gooners, etc.

18

[Solved] Radarr/Sonarr Remote Path Mapping

preview

Edit: Solved! Thanks a ton non_burglar.

Good day selfhost gang!

I’m redoing my media stack and I’ve come across a conundrum that I can’t solve on my own. I’ve got Docker containers on the same Debian machine for qbittorrent, sonarr, and radarr for automatic media downloading. The problem is, as you may have guessed from the title, I need to implement remote path mapping in sonarr/radarr because qbittorrent downloads to a different location than where my media files are hosted. My media is hosted on a network share, mounted at /home/marighost/gilashare/media (which is further broken into subfolders for media types). On this machine, however, I’ve got my qbittorrent downloading to a 2Tb SSD at /srv/data/torrents. Here are the volumes as presented in my compose.yml files:

qbittorrent:  
    volumes:  
      - /home/marighost/docker/seedbox/gluetunconfig:/gluetun  
      - /home/marighost/docker/seedbox/qbittorrentconfig:/config  
      - /srv/data/torrents:/data/torrents  

sonarr:  
    volumes:  
      - /home/marighost/docker/arrstack/sonarrconfig:/config  
      - /home/marighost/gilashare/media:/tv  
      - /srv/data/torrents:/data  

radarr:  
    volumes:  
      - /home/marighost/docker/arrstack/sonarrconfig:/config  
      - /home/marighost/gilashare/media/movies:/movies  
      - /srv/data/torrents:/data  

All three applications work mostly as expected (radarr can grab a release from an indexer, throws it into qbittorrent to download), but it does not move media into the /gilashare subfolder.

Obviously, I get the error in both sonarr and radarr that tells me: You are using docker; download client qBittorrent places downloads in /data/torrents but this directory does not appear to exist inside the container. Review your remote path mappings and container volume settings.

Here is how my remote path mapping looks in the settings of sonarr/radarr:

Host: qbittorrent  
Remote Path: /srv/data/torrents/  
Local Path: /data/  

I was messing around with radarr initially, and also tried appending /movies to the ends of the Remote and Local paths to see if it needed the full path, but that didn’t seem to work either. I’ve also read through Trash Guides, but the information there seems outdated (all references to images are just… not there). I’ve also read radarr’s docs, but I feel like I’m missing something, and they just link to Trash Guides anyway.

Thanks for reading, happy to dig into my configs or setup more if necessary.

Edit: Solved! Thanks a ton non_burglar.

Good day selfhost gang!

I’m redoing my media stack and I’ve come across a conundrum that I can’t solve on my own. I’ve got Docker containers on the same Debian machine for qbittorrent, sonarr, and radarr for automatic media downloading. The problem is, as you may have guessed from the title, I need to implement remote path mapping in sonarr/radarr because qbittorrent downloads to a different location than where my media files are hosted. My media is hosted on a network share, mounted at /home/marighost/gilashare/media (which is further broken into subfolders for media types). On this machine, however, I’ve got my qbittorrent downloading to a 2Tb SSD at /srv/data/torrents. Here are the volumes as presented in my compose.yml files:

qbittorrent:  
    volumes:  
      - /home/marighost/docker/seedbox/gluetunconfig:/gluetun  
      - /home/marighost/docker/seedbox/qbittorrentconfig:/config  
      - /srv/data/torrents:/data/torrents  

sonarr:  
    volumes:  
      - /home/marighost/docker/arrstack/sonarrconfig:/config  
      - /home/marighost/gilashare/media:/tv  
      - /srv/data/torrents:/data  

radarr:  
    volumes:  
      - /home/marighost/docker/arrstack/sonarrconfig:/config  
      - /home/marighost/gilashare/media/movies:/movies  
      - /srv/data/torrents:/data  

All three applications work mostly as expected (radarr can grab a release from an indexer, throws it into qbittorrent to download), but it does not move media into the /gilashare subfolder.

Obviously, I get the error in both sonarr and radarr that tells me: You are using docker; download client qBittorrent places downloads in /data/torrents but this directory does not appear to exist inside the container. Review your remote path mappings and container volume settings.

Here is how my remote path mapping looks in the settings of sonarr/radarr:

Host: qbittorrent  
Remote Path: /srv/data/torrents/  
Local Path: /data/  

I was messing around with radarr initially, and also tried appending /movies to the ends of the Remote and Local paths to see if it needed the full path, but that didn’t seem to work either. I’ve also read through Trash Guides, but the information there seems outdated (all references to images are just… not there). I’ve also read radarr’s docs, but I feel like I’m missing something, and they just link to Trash Guides anyway.

Thanks for reading, happy to dig into my configs or setup more if necessary.

24

[SOLVED] Caddy and multiple subdomains/reverse proxies

preview

Hoping the brain trust can help.

Single reverse proxy? No issues at all. If I add multiple per some instructions on the caddy forums, nothing works - the docker container just continually reboots with no errors in the log.

Caddyfile:

fmd.domain.ca {
    reverse_proxy 10.179.200.199:8080
}

llama.domain.ca {
    reverse_proxy 192.168.3.199:3000
)

With only the fmd entry, works fine.

Hoping the brain trust can help.

Single reverse proxy? No issues at all. If I add multiple per some instructions on the caddy forums, nothing works - the docker container just continually reboots with no errors in the log.

Caddyfile:

fmd.domain.ca {
    reverse_proxy 10.179.200.199:8080
}

llama.domain.ca {
    reverse_proxy 192.168.3.199:3000
)

With only the fmd entry, works fine.

-26

[AIP] Reloop: a self-hosted CRM that builds itself from your mailbox (AGPL, Docker)

preview

Hi, I’m the developer. Reloop is a CRM for small teams that reads your mailbox instead of waiting for you to type things in.

What it does:

  • Connects IMAP, Google Workspace or Microsoft 365 with read-only access. You pick how far back it reads.
  • Builds contacts, companies and a timeline per person.
  • Ranks former customers in a “win back” list. Each point of the score is shown as its own line, so the ranking is not a black box.
  • Drafts a follow-up, but never sends. The button opens your own mail client.
  • Since 0.22 it has a read-only MCP server, so ChatGPT, Claude or Claude Code can query your own instance.

Self-hosting:

curl -fsSL https://reloopcrm.com/install.sh | sh

Docker Compose with Postgres 17, API, web app and agent. amd64 and arm64. 8 GB RAM recommended, the agent is the heavy part. Without an AI key it runs as a normal CRM. Telemetry only exists if you set your own PostHog key.

AI disclosure: most of the code was written by AI coding agents that I directed and reviewed. The app itself uses an LLM only with a key you add.

Repo: github.com/reloopcrm/reloop

Feedback welcome, especially on the install and on what would make you trust software reading your mail.

Hi, I’m the developer. Reloop is a CRM for small teams that reads your mailbox instead of waiting for you to type things in.

What it does:

  • Connects IMAP, Google Workspace or Microsoft 365 with read-only access. You pick how far back it reads.
  • Builds contacts, companies and a timeline per person.
  • Ranks former customers in a “win back” list. Each point of the score is shown as its own line, so the ranking is not a black box.
  • Drafts a follow-up, but never sends. The button opens your own mail client.
  • Since 0.22 it has a read-only MCP server, so ChatGPT, Claude or Claude Code can query your own instance.

Self-hosting:

curl -fsSL https://reloopcrm.com/install.sh | sh

Docker Compose with Postgres 17, API, web app and agent. amd64 and arm64. 8 GB RAM recommended, the agent is the heavy part. Without an AI key it runs as a normal CRM. Telemetry only exists if you set your own PostHog key.

AI disclosure: most of the code was written by AI coding agents that I directed and reviewed. The app itself uses an LLM only with a key you add.

Repo: github.com/reloopcrm/reloop

Feedback welcome, especially on the install and on what would make you trust software reading your mail.

30

cheap remote kvm

preview

Does anyone have a solution for a cheap remote KVM? All the solutions I’ve found are around $100 and I don’t want to spend that much.

I might even settle for a portable KVM if it’s really cheap.

I have a laptop and an Android tablet that I could repurpose, if that helps at all.

Does anyone have a solution for a cheap remote KVM? All the solutions I’ve found are around $100 and I don’t want to spend that much.

I might even settle for a portable KVM if it’s really cheap.

I have a laptop and an Android tablet that I could repurpose, if that helps at all.

23

Netbird and Pihole

preview

I’ve got a modest homelab setup: a couple computers and a server with the pfsense router pointing all DNS queries to my pi hole on a Pi Zero W. The server exposes a few services via cloudflare tunnel. It’s been working pretty well.

Then I started setting up a server for a friend. It’s a big repurposed server board running proxmox and will eventually have a handful of VMs and LXCs, all on a network. It’s for his business, so I’m setting up Netbird on it for his half-dozen employees to be able to access it for fileshares and stuff. I’m connecting my Linux desktop, too, so I can do sysadmin for him.

The trouble is that domain name resolution is getting hairy between my pihole and his netbird. Since it’s not my network, I can’t add my pihole as a DNS resource, but I do need to keep a connection to his network.

Anyone have any wisdom to share as to how to handle this properly and still be able to use my pihole for DNS filtering in my house?

I’ve got a modest homelab setup: a couple computers and a server with the pfsense router pointing all DNS queries to my pi hole on a Pi Zero W. The server exposes a few services via cloudflare tunnel. It’s been working pretty well.

Then I started setting up a server for a friend. It’s a big repurposed server board running proxmox and will eventually have a handful of VMs and LXCs, all on a network. It’s for his business, so I’m setting up Netbird on it for his half-dozen employees to be able to access it for fileshares and stuff. I’m connecting my Linux desktop, too, so I can do sysadmin for him.

The trouble is that domain name resolution is getting hairy between my pihole and his netbird. Since it’s not my network, I can’t add my pihole as a DNS resource, but I do need to keep a connection to his network.

Anyone have any wisdom to share as to how to handle this properly and still be able to use my pihole for DNS filtering in my house?

34

[AIP] WebZFS - Modern Web Interface for ZFS Administration

preview

Been looking for a nice ZFS UI for my NAS after ditching TrueNAS for PBS. This is developed by a former TrueNAS/iXsystems employee and looking pretty good. Set it up and it’s pretty intuitive and helpful.

AI has been used for the UI but not the backend if I got it right (there’s a rant about that in the README).

Been looking for a nice ZFS UI for my NAS after ditching TrueNAS for PBS. This is developed by a former TrueNAS/iXsystems employee and looking pretty good. Set it up and it’s pretty intuitive and helpful.

AI has been used for the UI but not the backend if I got it right (there’s a rant about that in the README).

39

PSA: unattended-upgrades restarted nginx in the middle of my stream

preview

Small thing but it got me last week. I relay my stream to twitch through nginx-rtmp on a cheap VPS, and one morning unattended-upgrades pulled a new systemd, needrestart restarted nginx (and postgres, and everything else) and the stream dropped. OBS reconnected fast but twitch counted it as two streams.

default timer is 06:00 UTC + up to 1h random, which is exactly when I stream lol. moved it with a drop-in:

# systemctl edit apt-daily-upgrade.timer
[Timer]
OnCalendar=
OnCalendar=*-*-* 13:45:00 UTC
RandomizedDelaySec=15m

the empty OnCalendar= matters, without it you just add a second time and keep the old one.

anyone else got bitten by this?

Small thing but it got me last week. I relay my stream to twitch through nginx-rtmp on a cheap VPS, and one morning unattended-upgrades pulled a new systemd, needrestart restarted nginx (and postgres, and everything else) and the stream dropped. OBS reconnected fast but twitch counted it as two streams.

default timer is 06:00 UTC + up to 1h random, which is exactly when I stream lol. moved it with a drop-in:

# systemctl edit apt-daily-upgrade.timer
[Timer]
OnCalendar=
OnCalendar=*-*-* 13:45:00 UTC
RandomizedDelaySec=15m

the empty OnCalendar= matters, without it you just add a second time and keep the old one.

anyone else got bitten by this?

46

Can you recommend me a free domain service?

preview

I want to try out some things with a secondary domain. I’d prefer ethical and non-shady providers. I have a fixed IP and a VPS to go with it, so it’s really only about the domain. A subdomain of my current domain is not enough. A subdomain of the provider would be OK but not ideal.
Even paying very little is not an option right now - maybe later.

Doing some preliminary research, the keyword seems to be Dynamic DNS even if I have a static IP. There are providers but those that pop up easily do not instil much confidence.

I want to hear the Fediverse’s opinions! I haven’t found any recent posts about this.

TIA

I want to try out some things with a secondary domain. I’d prefer ethical and non-shady providers. I have a fixed IP and a VPS to go with it, so it’s really only about the domain. A subdomain of my current domain is not enough. A subdomain of the provider would be OK but not ideal.
Even paying very little is not an option right now - maybe later.

Doing some preliminary research, the keyword seems to be Dynamic DNS even if I have a static IP. There are providers but those that pop up easily do not instil much confidence.

I want to hear the Fediverse’s opinions! I haven’t found any recent posts about this.

TIA

35

I'm upgrading from an old pc to a NAS with OMV. What about RAID?

preview

So I’ve been running a home media server with the arr suite on it, but it only has about 2 tb storage. The storage has been full for a couple months now which has rendered my server useless. Now I’ve bought a used nas with 32 tb on 7 disks. I’ve been advised to set it up in RAID. I’ve learned how to do that in OMV and it’s setting that up now.

It takes a while though, so that leaves me some room to think. Why do we do this? I understand that it gives me more unified storage, but I also want to run other things on this machine. Like k8s to deploy my own applications and play around with devops stuff. Or even just regular file storage and photobackup next to the films and series. I thonk I would like to keep those things seperate though, is that still possible with RAID? Or is it better to put 5 disks in RAID and keep the other two for other stuff? I think my question is, how do you use RAID “practically”? How do you store different things on there? Do you compartmentalize the disks again, or do you just make different folders?

So I’ve been running a home media server with the arr suite on it, but it only has about 2 tb storage. The storage has been full for a couple months now which has rendered my server useless. Now I’ve bought a used nas with 32 tb on 7 disks. I’ve been advised to set it up in RAID. I’ve learned how to do that in OMV and it’s setting that up now.

It takes a while though, so that leaves me some room to think. Why do we do this? I understand that it gives me more unified storage, but I also want to run other things on this machine. Like k8s to deploy my own applications and play around with devops stuff. Or even just regular file storage and photobackup next to the films and series. I thonk I would like to keep those things seperate though, is that still possible with RAID? Or is it better to put 5 disks in RAID and keep the other two for other stuff? I think my question is, how do you use RAID “practically”? How do you store different things on there? Do you compartmentalize the disks again, or do you just make different folders?

55

[IAP] tinyfeed 1.5.1: the stability release!

GitHub Releases · TheBigRoomXXL/tinyfeed github.com
preview

tinyfeed is a CLI tool that generates a webpage from a collection of feeds. It’s dead simple: no database, no config file, just a CLI and some HTML.

This release focuses on making tinyfeed a more stable and deterministic program. A stable experience is always better for users, and it also allowed us to design a new snapshot-based test suite, so regressions are much less likely to slip through.

It also included a bunch of fixes related to the –order-by option that was added in the previous release and was, sadly, quite full of bugs 😬.

Main items on the menu today:

  • Stable ordering: Feeds and items now come out in the same order on every run.
  • Discoverable sorting: --help now lists the values –order-by accepts: publication-date, update-date, feed-name and author.
  • Lighter dependencies footprint: upgrad to gofeed v1.5, which removes 6 indirect dependencies.
  • Lots of refactoring and a brand new test suite.

And the cherry on top: tinyfeed now publishes an Atom feed, so you can follow its releases! WHY DIDN’T I THINK OF THIS EARLIER? A FEED FOR TINYFEED!

AI Disclosure:

  • Implementation - Pair: I like directing IA when I have a refectoring in mind. On new feature I prefere avoiding it. I find that using IA for feature mean I spends less time time using the software myself during dev and then I miss the bugs.
  • Testing - Hint: the new integration tests suit is 100% me, but the new unit tests are often generated by IA then reviewed / refactor by me.
  • Review - Pair: for contributions it’s 100% myself, I don’t have much MR but I take a lot of time to review them. For my own code, as im a solo maintenr, the IA is the only one giving me reivew.

tinyfeed is a CLI tool that generates a webpage from a collection of feeds. It’s dead simple: no database, no config file, just a CLI and some HTML.

This release focuses on making tinyfeed a more stable and deterministic program. A stable experience is always better for users, and it also allowed us to design a new snapshot-based test suite, so regressions are much less likely to slip through.

It also included a bunch of fixes related to the –order-by option that was added in the previous release and was, sadly, quite full of bugs 😬.

Main items on the menu today:

  • Stable ordering: Feeds and items now come out in the same order on every run.
  • Discoverable sorting: --help now lists the values –order-by accepts: publication-date, update-date, feed-name and author.
  • Lighter dependencies footprint: upgrad to gofeed v1.5, which removes 6 indirect dependencies.
  • Lots of refactoring and a brand new test suite.

And the cherry on top: tinyfeed now publishes an Atom feed, so you can follow its releases! WHY DIDN’T I THINK OF THIS EARLIER? A FEED FOR TINYFEED!

AI Disclosure:

  • Implementation - Pair: I like directing IA when I have a refectoring in mind. On new feature I prefere avoiding it. I find that using IA for feature mean I spends less time time using the software myself during dev and then I miss the bugs.
  • Testing - Hint: the new integration tests suit is 100% me, but the new unit tests are often generated by IA then reviewed / refactor by me.
  • Review - Pair: for contributions it’s 100% myself, I don’t have much MR but I take a lot of time to review them. For my own code, as im a solo maintenr, the IA is the only one giving me reivew.
35

Folder/file name structure for eBooks

preview

Howdy self-hosters!

I’ve got a huge pile of eBooks flying around here: poetry, scientific literature, magazines, comics, teaching materials, RPG rule books - it’s a wildly mixed bag. I’m still looking for a self-hosted solution to make it more accessible and get an overview of what I actually have. (Calibre looks promising!)

My question here is not about software, though, but about organisation at the filesystem level: what folder and file name structure have you established for your eBooks? Do you just dump everything into one single folder and let the software do the rest? Do you create sections by genre (“Crime”, “Fantasy”, “Young adult”, like the libraries do it) or application (job-related, hobby-related, cooking, …)? Does it matter at all?

Howdy self-hosters!

I’ve got a huge pile of eBooks flying around here: poetry, scientific literature, magazines, comics, teaching materials, RPG rule books - it’s a wildly mixed bag. I’m still looking for a self-hosted solution to make it more accessible and get an overview of what I actually have. (Calibre looks promising!)

My question here is not about software, though, but about organisation at the filesystem level: what folder and file name structure have you established for your eBooks? Do you just dump everything into one single folder and let the software do the rest? Do you create sections by genre (“Crime”, “Fantasy”, “Young adult”, like the libraries do it) or application (job-related, hobby-related, cooking, …)? Does it matter at all?

19

Help with Netbird Networks access to home resources

preview

I have a GLInet travel router that is a peer connected to my Netbird network, I’d like to have any client that connects to my router to have access to the peers the travel router has access to, but still be able to access the regular internet that is being passed through by my router. Is this possible?

Ultimately I’m trying to get the TV at my girlfriend’s place to be able to access my Jellyfin instance over Netbird. According to the Netbird documentation, exit nodes and routes are depricated and should not be used, but I’m struggling hard with getting it to work with Networks.

  • I’ve created a group with my GLInet router, my Jellyfin instance, and my home server (Which hosts Jellyfin) all in the group called “Travel”.
  • I’ve created a Network called “Home” in Netbird
  • Within the “Home” network I’ve created a resource called “Jellyfin” with the IP address set to it’s Netbird IP address, and an access control group where all members of the “Travel” group have access to all ports of the “Jellyfin” resource.
  • I’ve created two routing peers within the network, one is the GLInet router and the other is my home server. Both of these are Netbird peers with Netbird installed on them.

I am very confused overall, and could use a hand if anyone wanted to walk me through what I’m doing wrong.

I have a GLInet travel router that is a peer connected to my Netbird network, I’d like to have any client that connects to my router to have access to the peers the travel router has access to, but still be able to access the regular internet that is being passed through by my router. Is this possible?

Ultimately I’m trying to get the TV at my girlfriend’s place to be able to access my Jellyfin instance over Netbird. According to the Netbird documentation, exit nodes and routes are depricated and should not be used, but I’m struggling hard with getting it to work with Networks.

  • I’ve created a group with my GLInet router, my Jellyfin instance, and my home server (Which hosts Jellyfin) all in the group called “Travel”.
  • I’ve created a Network called “Home” in Netbird
  • Within the “Home” network I’ve created a resource called “Jellyfin” with the IP address set to it’s Netbird IP address, and an access control group where all members of the “Travel” group have access to all ports of the “Jellyfin” resource.
  • I’ve created two routing peers within the network, one is the GLInet router and the other is my home server. Both of these are Netbird peers with Netbird installed on them.

I am very confused overall, and could use a hand if anyone wanted to walk me through what I’m doing wrong.

15

GPU dri/by-path for Docker Compose how?

preview

My Jellyfin config has some YAML like this.

# Sometimes work  
devices:  
- /dev/dri/renderD128:/dev/dri/renderD128  
- /dev/dri/card0:/dev/dri/card0  

Occasionally when I reboot my server, I end up with no card0 and instead I have /dev/dri/card1. This breaks my Jellyfin server…

I noticed there is also dri/by-path, which I’m guessing is a more stable path than a card index… however, doing this doesn’t work for me.

# No work  
devices:  
- /dev/dri/by-path/pci-0000:00:02.0-render:/dev/dri/renderD128  
- /dev/dri/by-path/pci-0000:00:02.0-card:/dev/dri/card0  

I tried with and without quotes and the result is the same. Error to start. I’m guessing Docker is failing to correctly parse the string because of the colons in pci-0000:00:02.0-card…

How’s everyone else handling this problem? I’ve been updating my compose.yaml every time the index switches, card0 to card1 to card0… but this is getting annoying now. (This is a worse problem now that I’m running Fedora IoT!)

Any thoughts?

My Jellyfin config has some YAML like this.

# Sometimes work  
devices:  
- /dev/dri/renderD128:/dev/dri/renderD128  
- /dev/dri/card0:/dev/dri/card0  

Occasionally when I reboot my server, I end up with no card0 and instead I have /dev/dri/card1. This breaks my Jellyfin server…

I noticed there is also dri/by-path, which I’m guessing is a more stable path than a card index… however, doing this doesn’t work for me.

# No work  
devices:  
- /dev/dri/by-path/pci-0000:00:02.0-render:/dev/dri/renderD128  
- /dev/dri/by-path/pci-0000:00:02.0-card:/dev/dri/card0  

I tried with and without quotes and the result is the same. Error to start. I’m guessing Docker is failing to correctly parse the string because of the colons in pci-0000:00:02.0-card…

How’s everyone else handling this problem? I’ve been updating my compose.yaml every time the index switches, card0 to card1 to card0… but this is getting annoying now. (This is a worse problem now that I’m running Fedora IoT!)

Any thoughts?

71

'My own piece of the Internet: Part 1', a blog post discussing the process of getting an ARIN ASN and a /24 IPv4 block for their home setup

Gonzalo Escarrá's Blog My own piece of the Internet: Part 1 blog.escarra.org
preview

From the blog post

obtaining an ARIN ASN and a portable /24 IPv4 block for a home networkI’ve wanted my own Internet address space for a long time. Not “my ISP gave me some statics” space which I’ve had for well over a decade now, but a block that belongs to me, announced into the global routing table by equipment I own, and ideally sitting in the rack in my house. Type the prefix into bgp.tools and see my own ASN sitting there as the origin. How cool! (ok, cool for some, but not most)

From the blog post

obtaining an ARIN ASN and a portable /24 IPv4 block for a home networkI’ve wanted my own Internet address space for a long time. Not “my ISP gave me some statics” space which I’ve had for well over a decade now, but a block that belongs to me, announced into the global routing table by equipment I own, and ideally sitting in the rack in my house. Type the prefix into bgp.tools and see my own ASN sitting there as the origin. How cool! (ok, cool for some, but not most)

28

Backup

preview

I view my photos using immich, historically I used photo prism with photosync to back them up.

My photos are stored on my nas (omv), immich is on my android and I access it through wireguard.

My nas is raidz1 with 4 drives, so it is relatively resilient. The photos are backed up to a separate external hard drive too using borgbackup.

What i’d like to do is the following: Back up my photos externally - at a family members house with a pi zero connected to a hard drive, which would be connected to their WiFi or via an ethernet cable to the router.

Ideally, the first backup should be done at home before, I don’t want to backup 100’s GB over the internet.

Would a simple wireguard tunnel be ok here? Should a different method be used?

Has anyone done such a thing? I read a lot about vps, but I don’t what the ongoing commitment in cost when I have a free option.

One of my worries is that I don’t want a compromised device on the family member’s network to compromise mine, would the wireguard tunnel be enough to secure it? Would WiFi be more secure than ethernet cable?

I view my photos using immich, historically I used photo prism with photosync to back them up.

My photos are stored on my nas (omv), immich is on my android and I access it through wireguard.

My nas is raidz1 with 4 drives, so it is relatively resilient. The photos are backed up to a separate external hard drive too using borgbackup.

What i’d like to do is the following: Back up my photos externally - at a family members house with a pi zero connected to a hard drive, which would be connected to their WiFi or via an ethernet cable to the router.

Ideally, the first backup should be done at home before, I don’t want to backup 100’s GB over the internet.

Would a simple wireguard tunnel be ok here? Should a different method be used?

Has anyone done such a thing? I read a lot about vps, but I don’t what the ongoing commitment in cost when I have a free option.

One of my worries is that I don’t want a compromised device on the family member’s network to compromise mine, would the wireguard tunnel be enough to secure it? Would WiFi be more secure than ethernet cable?

152

Is anyone self-hosting their own AI?

preview

Does anyone run their own AI service?

If so,

Which one?

What hardware?

Do you have a way to access it from outside or from a phone?

And, if you’re open to sharing, what do you do with it?

I know AI-enabling hardware is insanely expensive right now, but some of the smaller (<32GB) more efficient models (think deepseek- or qwen-based) don’t absolutely need specialized hardware to run and perform well enough for many purposes.

I am so aware of the absolutely abysmal privacy practices of basically all AI companies that I’m not comfortable using them to help me with web searches and the such, because there’s no way that all this information won’t be going into a profile. This makes me want to have my own “helper” service.

Does anyone run their own AI service?

If so,

Which one?

What hardware?

Do you have a way to access it from outside or from a phone?

And, if you’re open to sharing, what do you do with it?

I know AI-enabling hardware is insanely expensive right now, but some of the smaller (<32GB) more efficient models (think deepseek- or qwen-based) don’t absolutely need specialized hardware to run and perform well enough for many purposes.

I am so aware of the absolutely abysmal privacy practices of basically all AI companies that I’m not comfortable using them to help me with web searches and the such, because there’s no way that all this information won’t be going into a profile. This makes me want to have my own “helper” service.

108

What else should I do for my home server security?

preview

So I’m relatively new to server hosting. (But experienced with remote web hosting from providers.) After about a year of fighting to run Nextcloud on a hosting company and finally losing everything from the cPanel vulnerability a few months ago, I broke down and launched my own home server. I’m looking for advice to maximize security.

CachyOS Linux with ClamAV Docker via Portainer Nextcloud (E2EE), Immich, DocuSeal and NoteTrace Routed through a domain and subdimains via Zero Trust Cloudflare Tunnel and Cloudflare DNS.

I have regular redundant backups both onsite and offsite with Duplicati.

My passwords for OS and all admin dashboards are extremely long, unique, random strings managed by Bitwarden.

I know just enough about this stuff to get into trouble. What else should I setup? I have little to no monitoring going on and don’t know much about that direction.

So I’m relatively new to server hosting. (But experienced with remote web hosting from providers.) After about a year of fighting to run Nextcloud on a hosting company and finally losing everything from the cPanel vulnerability a few months ago, I broke down and launched my own home server. I’m looking for advice to maximize security.

CachyOS Linux with ClamAV Docker via Portainer Nextcloud (E2EE), Immich, DocuSeal and NoteTrace Routed through a domain and subdimains via Zero Trust Cloudflare Tunnel and Cloudflare DNS.

I have regular redundant backups both onsite and offsite with Duplicati.

My passwords for OS and all admin dashboards are extremely long, unique, random strings managed by Bitwarden.

I know just enough about this stuff to get into trouble. What else should I setup? I have little to no monitoring going on and don’t know much about that direction.

37

HomeAssist Kiosks

preview

I’m looking for advice on setting up kiosks for a smart home.

I’d like to have 2-3 touch screens setup to control devices from.

I’ve decided on the following:

  • preferably plug in devices/screens (with no battery)
  • e-ink screens would be preferable but not a must
  • cost is a factor

Anyone had success in setting up something similar?

I’m looking for advice on setting up kiosks for a smart home.

I’d like to have 2-3 touch screens setup to control devices from.

I’ve decided on the following:

  • preferably plug in devices/screens (with no battery)
  • e-ink screens would be preferable but not a must
  • cost is a factor

Anyone had success in setting up something similar?

48

How long do you keep your backups?

preview

Right now I’m creating a nightly restic snapshot of my server on borgbase and on some spare drives, and keeping the snapshots forever. There’s a ton of daily churn in my data though, between databases and logs and other files that don’t constitute ‘real data’ but might be needed to restore the app.

Right now I’m creating a nightly restic snapshot of my server on borgbase and on some spare drives, and keeping the snapshots forever. There’s a ton of daily churn in my data though, between databases and logs and other files that don’t constitute ‘real data’ but might be needed to restore the app.

7

Wattroom for bike training

GitHub GitHub - natrontech/wattroom: Train together, not alone. Collaborative indoor cycling: structured workouts, FTMS trainer control, crews with voice channels, live metrics, camera and a synced jukebox. github.com
preview

I saw this and installed it but the WebBluetooth doesn’t work regardless of the browser I used on Android. Can someone tell me if it’s possible to use this on your phone? I can’t move my desktop to where the bike is. Also I’ve never used Chrome on my phone and I feel raped sybernetically for being used like that to install that piece of shit data hording browser.

I saw this and installed it but the WebBluetooth doesn’t work regardless of the browser I used on Android. Can someone tell me if it’s possible to use this on your phone? I can’t move my desktop to where the bike is. Also I’ve never used Chrome on my phone and I feel raped sybernetically for being used like that to install that piece of shit data hording browser.

41

Improving security by separating server duties

preview

Hi selfhosters,

I’m considering improving the security of my homelab/server setup and I’d love to hear some more experienced opinions on my plans.

Like most self-hosters, I started out with a single-server setup: I cobbled together some old hardware, slapped on a NAS-style OS, bolted on Docker and started experimenting. As things go, my setup has grown a lot over the years. I’m running a lot on that machine - NAS capability, local DNS via pihole, a Wireguard VPN endpoint (so I can tunnel into the LAN from outside), many private services (accessible only from inside the LAN), but also some public ones accessible from the open web.

Although I’ve tried to do my homework in terms of security (various hardening measures for SSH and Docker, reverse proxy, proper SSL termination for everything), I realise I’ve put a lot of eggs in that one basket and I might want to introduce some network segmentation.

I don’t think I’m ready for getting a hardware firewall or a router/switch capable of VLANs just yet. What I’m considering, though, is putting my services available from the open internet on a separate server inside my network - a little Pi 5 I’ve still got lying around. My idea is that if a container on that is breached and if an attacker manages to escape the container to gain root access, compromise would be limited to the few of publicly available services I have on the Pi, thus limiting blast radius. If combined with (software) firewall rules that make the Pi a ‘pariah node’ (dropping all incoming connection efforts), lateral movement to other devices on the network could be contained.

An alternative would be to finally make the move to Proxmox. Everything would still be running on one physical machine, but I’d create several VMs: there would be one for private, one for publicly available services, and I’d firewall the latter off rigorously so it cannot connect to anything else.

I’d like your opinion on these plans. Which one would you recommend? Are they both sound, or is there anything I haven’t considered? Are there other (easy, non-hardware intense) things I could do to improve security even more?

Hi selfhosters,

I’m considering improving the security of my homelab/server setup and I’d love to hear some more experienced opinions on my plans.

Like most self-hosters, I started out with a single-server setup: I cobbled together some old hardware, slapped on a NAS-style OS, bolted on Docker and started experimenting. As things go, my setup has grown a lot over the years. I’m running a lot on that machine - NAS capability, local DNS via pihole, a Wireguard VPN endpoint (so I can tunnel into the LAN from outside), many private services (accessible only from inside the LAN), but also some public ones accessible from the open web.

Although I’ve tried to do my homework in terms of security (various hardening measures for SSH and Docker, reverse proxy, proper SSL termination for everything), I realise I’ve put a lot of eggs in that one basket and I might want to introduce some network segmentation.

I don’t think I’m ready for getting a hardware firewall or a router/switch capable of VLANs just yet. What I’m considering, though, is putting my services available from the open internet on a separate server inside my network - a little Pi 5 I’ve still got lying around. My idea is that if a container on that is breached and if an attacker manages to escape the container to gain root access, compromise would be limited to the few of publicly available services I have on the Pi, thus limiting blast radius. If combined with (software) firewall rules that make the Pi a ‘pariah node’ (dropping all incoming connection efforts), lateral movement to other devices on the network could be contained.

An alternative would be to finally make the move to Proxmox. Everything would still be running on one physical machine, but I’d create several VMs: there would be one for private, one for publicly available services, and I’d firewall the latter off rigorously so it cannot connect to anything else.

I’d like your opinion on these plans. Which one would you recommend? Are they both sound, or is there anything I haven’t considered? Are there other (easy, non-hardware intense) things I could do to improve security even more?

163

Finally escaping Windows, ironically thanks to AI

preview

I’ve had several occasions in my life where I’ve tried to switch to Linux only to run into hardware incompatibilities that would have taken me man-months to solve. Thanks to GPT my primary personal PC, home media and project server is now running Linux and fully functional. I’ve got Jellyfin, NextCloud, Navidrome and Joplin all happy. I’ve also been able to wipe an obsolete MacOS off my 2017-era Macbook Pro and it’s happily running Mint.

There’s a deep irony in Windows being enshittified by AI, while the same AI makes Linux about a billion times more usable. Now I have a leftover Windows micro-PC running a local LLM model, so I can be less dependent on GPT, as long as time isn’t a factor…

I’ve had several occasions in my life where I’ve tried to switch to Linux only to run into hardware incompatibilities that would have taken me man-months to solve. Thanks to GPT my primary personal PC, home media and project server is now running Linux and fully functional. I’ve got Jellyfin, NextCloud, Navidrome and Joplin all happy. I’ve also been able to wipe an obsolete MacOS off my 2017-era Macbook Pro and it’s happily running Mint.

There’s a deep irony in Windows being enshittified by AI, while the same AI makes Linux about a billion times more usable. Now I have a leftover Windows micro-PC running a local LLM model, so I can be less dependent on GPT, as long as time isn’t a factor…

571

Home server

preview

Built when Netflix first took away password sharing. Thanks for the heads up Netflix! Would not want to start this project now 😂

Edit: adding temp and specs

I9 11900k Z590I Vision D ITX mobo 64gb memory The 16x lane is bifurcated to run: Intel a310 & Lsi 9305 16e into a hp dl380

Drives are 10x 14 tb wdd in raidz1, 6 256gb sata ssds in a pair of 3 wide mirrors for special vdev metadata, a 512gb nvme for read cache, and a 64gb optane for slog

Built when Netflix first took away password sharing. Thanks for the heads up Netflix! Would not want to start this project now 😂

Edit: adding temp and specs

I9 11900k Z590I Vision D ITX mobo 64gb memory The 16x lane is bifurcated to run: Intel a310 & Lsi 9305 16e into a hp dl380

Drives are 10x 14 tb wdd in raidz1, 6 256gb sata ssds in a pair of 3 wide mirrors for special vdev metadata, a 512gb nvme for read cache, and a 64gb optane for slog

142

My lil server (RasPi4+powered USB hub with 2 SSDs)

preview

I run DietPI on one of my SSDs. SSDs are inside USB adapters connected to USB hub. Recently I started from scratch.

I have tested already:

  • dockerized Navidrome music library with Tempus android APP via Tailscale. It needs some tweaks I guess for losing connection or music streaming stutters though.
  • Moonlight + Sunshine game streaming. Although I have not figured out how to handle problematic wireless bluetooth ps4 connection in RetroArch.

Starting from scratch so I guess I will stick to basics for now

  • SSH
  • dockerized Pi-Hole

Then maybe will gry to play with dockerized

  • Forgejo
  • Nextcloud
  • Obsidian (self-hosting?) abilities
  • Teamspeak and move from there.

Im more interested in development now, less in media server stuff.

Any upgrade recommendations are welcome. I thought about Thin client like wase, but will swe. I still have some empty slots for 2 more ssds ;P

I run DietPI on one of my SSDs. SSDs are inside USB adapters connected to USB hub. Recently I started from scratch.

I have tested already:

  • dockerized Navidrome music library with Tempus android APP via Tailscale. It needs some tweaks I guess for losing connection or music streaming stutters though.
  • Moonlight + Sunshine game streaming. Although I have not figured out how to handle problematic wireless bluetooth ps4 connection in RetroArch.

Starting from scratch so I guess I will stick to basics for now

  • SSH
  • dockerized Pi-Hole

Then maybe will gry to play with dockerized

  • Forgejo
  • Nextcloud
  • Obsidian (self-hosting?) abilities
  • Teamspeak and move from there.

Im more interested in development now, less in media server stuff.

Any upgrade recommendations are welcome. I thought about Thin client like wase, but will swe. I still have some empty slots for 2 more ssds ;P

49

Self hosting as a gift

preview

Hello selfhosters! I read here a lot, longing for the day to get back into it. I have a little plan to do so I have access to some thin clients with decent enough hardware to run a small file server. I’d like to set one up for myself and as a gift to my partner

I keep getting lost in the jungle of options and incompatibilities so I was hoping to enlist some help in selecting a stack

I’m looking for open source software in the base, maybe truenas but I prefer debian and my partner uses LMDE already.

  • On top of that should be a simple file server for backups/network storage. Some interface for managing storeage would be huge too
  • Wireguard with an interface
  • Having a simple container interface for future features like homeassistant and jellyfin et al is great too.

Open source and free is a must and I want to avoid risk for future lockins and freemium/ripe for enshitification-software as far as possible. And no or at least responsible llm use if that even exists anymore. Happy to shed featues to acheive that in combination with longevity. Is plain debian or even ubuntu the way to go or are there more optimised distros? Or am i just describing fedora with cockpit? Im not lost in linux itself, just overwhelmed with the endless options hoping someone can help me make a plan!

Would you make someone you love a home server?

Hello selfhosters! I read here a lot, longing for the day to get back into it. I have a little plan to do so I have access to some thin clients with decent enough hardware to run a small file server. I’d like to set one up for myself and as a gift to my partner

I keep getting lost in the jungle of options and incompatibilities so I was hoping to enlist some help in selecting a stack

I’m looking for open source software in the base, maybe truenas but I prefer debian and my partner uses LMDE already.

  • On top of that should be a simple file server for backups/network storage. Some interface for managing storeage would be huge too
  • Wireguard with an interface
  • Having a simple container interface for future features like homeassistant and jellyfin et al is great too.

Open source and free is a must and I want to avoid risk for future lockins and freemium/ripe for enshitification-software as far as possible. And no or at least responsible llm use if that even exists anymore. Happy to shed featues to acheive that in combination with longevity. Is plain debian or even ubuntu the way to go or are there more optimised distros? Or am i just describing fedora with cockpit? Im not lost in linux itself, just overwhelmed with the endless options hoping someone can help me make a plan!

Would you make someone you love a home server?

35

I’m looking for the best remote access solution that works for me

preview

Hello,

I stopped using TeamViewer a very long time ago. Since then, I’ve been using DWService and I’m quite happy with it. I like the fact that I can access to the terminal without launching any graphical component on the machine (which is handy for my headless servers, and allows me not to open SSH port). I also like the fact that it works even if my VPN server is down, as I can access remote servers that aren’t accessible via the web when they’re no longer reachable through my VPN.

However, there are a few issues that are bothering me more and more. I’ve switched my mum’s computer to Fedora and, whilst it worked perfectly under Windows, remote control is very buggy under fedora because the connection isn’t great. Furthermore, it’s impossible to access the screen if the computer isn’t unlocked.

So, I’d really like to find a solution that meets my criteria:

  • Accessible via a web browser
  • Ideally open-source
  • Free
  • Ideally self-hostable, but I’m worried that if, for example, my reverse proxy or VPN crashes, I won’t be able to access my machines to debug them?
  • Allows access to servers without a monitor
  • Allows access to computers running Fedora without the need for prior unlocking

Thanks!

Hello,

I stopped using TeamViewer a very long time ago. Since then, I’ve been using DWService and I’m quite happy with it. I like the fact that I can access to the terminal without launching any graphical component on the machine (which is handy for my headless servers, and allows me not to open SSH port). I also like the fact that it works even if my VPN server is down, as I can access remote servers that aren’t accessible via the web when they’re no longer reachable through my VPN.

However, there are a few issues that are bothering me more and more. I’ve switched my mum’s computer to Fedora and, whilst it worked perfectly under Windows, remote control is very buggy under fedora because the connection isn’t great. Furthermore, it’s impossible to access the screen if the computer isn’t unlocked.

So, I’d really like to find a solution that meets my criteria:

  • Accessible via a web browser
  • Ideally open-source
  • Free
  • Ideally self-hostable, but I’m worried that if, for example, my reverse proxy or VPN crashes, I won’t be able to access my machines to debug them?
  • Allows access to servers without a monitor
  • Allows access to computers running Fedora without the need for prior unlocking

Thanks!

-16

Tab Manager & Session Saver – local tab session backup for Firefox

preview

Extension for saving and restoring named tab sessions in Firefox — relevant for anyone with complex browsing setups or multiple projects.

What it does:

  • Save named snapshots of current tabs (‘work-project’, ‘research’, ‘current-bug’)
  • Restore any saved session without losing current tabs
  • Export/import sessions as plain JSON

All storage is local — no accounts, no cloud, no telemetry. The JSON is human-readable so you can inspect, back up, or version control your sessions.

I use it mostly for project context switching — save the tabs for one context before switching to another, restore when returning.

Extension for saving and restoring named tab sessions in Firefox — relevant for anyone with complex browsing setups or multiple projects.

What it does:

  • Save named snapshots of current tabs (‘work-project’, ‘research’, ‘current-bug’)
  • Restore any saved session without losing current tabs
  • Export/import sessions as plain JSON

All storage is local — no accounts, no cloud, no telemetry. The JSON is human-readable so you can inspect, back up, or version control your sessions.

I use it mostly for project context switching — save the tabs for one context before switching to another, restore when returning.

56

ISP won't set up static IP - Services keep going down

preview

I called my ISP today to set up a static IP because my services keep being taken down from my IP changing. I either need a business plan or a 5-10gbps residential plan to do that…

I cannot get on the business plan without a valid Business ID so that’s not an option. And no way in hell I’m paying for those speeds.

Anyone have any alternatives to stop this from happening so often? My public IP seems to change regularly every 2-3 weeks and it’s getting pretty frustrating having to reconfigure things. I only have 2 ISPs in my area and they are both terrible. I assume I’m out of luck, but maybe there’s some tech magic I’m unaware of.

I called my ISP today to set up a static IP because my services keep being taken down from my IP changing. I either need a business plan or a 5-10gbps residential plan to do that…

I cannot get on the business plan without a valid Business ID so that’s not an option. And no way in hell I’m paying for those speeds.

Anyone have any alternatives to stop this from happening so often? My public IP seems to change regularly every 2-3 weeks and it’s getting pretty frustrating having to reconfigure things. I only have 2 ISPs in my area and they are both terrible. I assume I’m out of luck, but maybe there’s some tech magic I’m unaware of.

14

[CBH] Ideon 1.0: a canvas that has become a Git artifact

preview

Hi folks!

Ten months ago I started Ideon as a side project. This week it reaches v1.0, with 640 stars on GitHub. Thanks to those who have followed along from the beginning, this really means a lot.

What’s new in 1.0: the canvas has become a Git artifact.

A project now syncs to a Git repository. The structure, notes and block positions are version-controlled as files inside an .ideon/ directory, with a manifest and SHA-256 checksums. The rest of your repo stays untouched.

In short: you edit the canvas, you push, and the project history sits in Git like code. Every change shows up as a file diff.

The rest of 1.0:

  • Database block: typed columns, search and aggregation, all on the canvas.
  • Mermaid block: live rendering, editable source, PNG export.
  • Chained audit trail and sealed records: every state change is traceable and verifiable.

What’s next:

  • A plugin API, so the community can build its own block types.
  • A desktop app. It’s Docker-only for now. I’m solo and can’t test every platform properly. That comes if the growth justifies it.

Where this goes next: the canvas stops being a map and starts being a system. A graph that remembers everything, connects the dots, and eventually reacts on its own 👀.

If you want to try it: a plain docker run works for the SQLite build, docker-compose if you prefer Postgres. The repo-sync guide is in the documentation.

The project: github.com/3xpyth0n/ideon

Hi folks!

Ten months ago I started Ideon as a side project. This week it reaches v1.0, with 640 stars on GitHub. Thanks to those who have followed along from the beginning, this really means a lot.

What’s new in 1.0: the canvas has become a Git artifact.

A project now syncs to a Git repository. The structure, notes and block positions are version-controlled as files inside an .ideon/ directory, with a manifest and SHA-256 checksums. The rest of your repo stays untouched.

In short: you edit the canvas, you push, and the project history sits in Git like code. Every change shows up as a file diff.

The rest of 1.0:

  • Database block: typed columns, search and aggregation, all on the canvas.
  • Mermaid block: live rendering, editable source, PNG export.
  • Chained audit trail and sealed records: every state change is traceable and verifiable.

What’s next:

  • A plugin API, so the community can build its own block types.
  • A desktop app. It’s Docker-only for now. I’m solo and can’t test every platform properly. That comes if the growth justifies it.

Where this goes next: the canvas stops being a map and starts being a system. A graph that remembers everything, connects the dots, and eventually reacts on its own 👀.

If you want to try it: a plain docker run works for the SQLite build, docker-compose if you prefer Postgres. The repo-sync guide is in the documentation.

The project: github.com/3xpyth0n/ideon

40

Which domain registry or specific TLD do you prefer, and why?

preview

(Not the registrar. E.g., .com TLD is operated by Verisign.)

I wanted a domain for my e-mail. Sounds simple.
But I am a major overthinker, so over the period of past 2 weeks (damn!) I’ve been deciding between e-mail providers, choosing potential domain names, choosing a TLD (round 1 - by price and looks), looking through registrars, deciding whom to choose for nameservers, and now I am comparing registry operators (TLD round 2).

There are differences, such as who can register a domain, restrictions on domain privacy, and some TLDs like .dev are on HSTS preload list.


Currently, I am mostly looking at EURid with .eu TLD.
What piqued my interest is being able to transfer a domain directly through the registry. So, if a registrar locked me out or shut down, I could still transfer my domain.
From EURid:

From your registrar: Your first point of contact should be your registrar.
From the My .eu platform: You can also issue the transfer authorization code yourself on the My.eu account.
From EURid: This should be done only as a last resort. After proving you are the holder of the domain and cannot obtain an authorization code from your current registrar or via the My .eu platform, you can request an emergency transfer authorization code directly from EURid

Alright.
But I must wonder, how do I prove my ownership?

First sign that this may not be that secure is right on the same page:

Print the form, sign it and send it to us via email at authcoderequest@eurid.eu. If the requester is an organization, please print the form on company letterhead paper or add a company stamp.

OK, the form. Here’s the request page. Fields are domain name, first name, last name, job title, email address, and reason for the request.

I generated the form to see what’s there. So… is this the security?:

I hereby confirm that I am the rightful holder of the domain name mentioned above, or that I have obtained the legal power to represent the holder, and that all data provided is true and correct. I am fully aware that making a false statement in this matter is considered fraud and may have judicial consequences.

I mean, it doesn’t even sound like they collect your signature during the verification, not that a picture provides much security. This is one of the methods:

You can receive a small bank payment (less than one euro) from EURid, which includes a 10-character code that starts and ends with “EU”, with six digits in between (EUXXXXXXEU). After receiving this payment, the domain holder logs in to the My .eu platform and enters the code to verify their registration data.

Either way, it seems this might be mitigated by registry lock:

The Registrar Lock Service will also protect your domain name from accidental changes, but the difference from Registry Lock is that a locked domain name can still be transferred with an authorisation code sent to your My .eu account or an emergency authorisation code received from EURid.

Enough about safety. What about privacy?
The T&C [PDF] prohibits the domain having info of the registrar:

Your registration data must be your own data and must not be that of your Registrar, or of a proxy holder or a representative of a natural person or legal entity who does not meet the eligibility criteria. You may at any time check your registration data via the web-based WHOIS or the secure registrant portal available on our Website.

But it gets a bit better in the WHOIS policy [PDF]:

When the Registrant is a natural person the Registry publishes the Registrant´s e-mail address and language.


Alright. I hope there’s someone else who took a similarly overcomplicated route in choosing a domain name.

(Not the registrar. E.g., .com TLD is operated by Verisign.)

I wanted a domain for my e-mail. Sounds simple.
But I am a major overthinker, so over the period of past 2 weeks (damn!) I’ve been deciding between e-mail providers, choosing potential domain names, choosing a TLD (round 1 - by price and looks), looking through registrars, deciding whom to choose for nameservers, and now I am comparing registry operators (TLD round 2).

There are differences, such as who can register a domain, restrictions on domain privacy, and some TLDs like .dev are on HSTS preload list.


Currently, I am mostly looking at EURid with .eu TLD.
What piqued my interest is being able to transfer a domain directly through the registry. So, if a registrar locked me out or shut down, I could still transfer my domain.
From EURid:

From your registrar: Your first point of contact should be your registrar.
From the My .eu platform: You can also issue the transfer authorization code yourself on the My.eu account.
From EURid: This should be done only as a last resort. After proving you are the holder of the domain and cannot obtain an authorization code from your current registrar or via the My .eu platform, you can request an emergency transfer authorization code directly from EURid

Alright.
But I must wonder, how do I prove my ownership?

First sign that this may not be that secure is right on the same page:

Print the form, sign it and send it to us via email at authcoderequest@eurid.eu. If the requester is an organization, please print the form on company letterhead paper or add a company stamp.

OK, the form. Here’s the request page. Fields are domain name, first name, last name, job title, email address, and reason for the request.

I generated the form to see what’s there. So… is this the security?:

I hereby confirm that I am the rightful holder of the domain name mentioned above, or that I have obtained the legal power to represent the holder, and that all data provided is true and correct. I am fully aware that making a false statement in this matter is considered fraud and may have judicial consequences.

I mean, it doesn’t even sound like they collect your signature during the verification, not that a picture provides much security. This is one of the methods:

You can receive a small bank payment (less than one euro) from EURid, which includes a 10-character code that starts and ends with “EU”, with six digits in between (EUXXXXXXEU). After receiving this payment, the domain holder logs in to the My .eu platform and enters the code to verify their registration data.

Either way, it seems this might be mitigated by registry lock:

The Registrar Lock Service will also protect your domain name from accidental changes, but the difference from Registry Lock is that a locked domain name can still be transferred with an authorisation code sent to your My .eu account or an emergency authorisation code received from EURid.

Enough about safety. What about privacy?
The T&C [PDF] prohibits the domain having info of the registrar:

Your registration data must be your own data and must not be that of your Registrar, or of a proxy holder or a representative of a natural person or legal entity who does not meet the eligibility criteria. You may at any time check your registration data via the web-based WHOIS or the secure registrant portal available on our Website.

But it gets a bit better in the WHOIS policy [PDF]:

When the Registrant is a natural person the Registry publishes the Registrant´s e-mail address and language.


Alright. I hope there’s someone else who took a similarly overcomplicated route in choosing a domain name.

10

Anyone running Goatcounter? Hyper minimal analytics.

preview

cross-posted from: discuss.online/post/46782639

Goatcounter is so minimal that it takes a minute to wrap the brain around, but the good news is that it tracks all access anonymously, which is all I ever wanted (and why I simply have never made real use of Matomo over the years). It is the digital equivalent of a step counter, and acts like those counting clickers they use at concerts to track number of attendants.

It is good to confirm people are actually able to access tooling. That is all I need.

cross-posted from: discuss.online/post/46782639

Goatcounter is so minimal that it takes a minute to wrap the brain around, but the good news is that it tracks all access anonymously, which is all I ever wanted (and why I simply have never made real use of Matomo over the years). It is the digital equivalent of a step counter, and acts like those counting clickers they use at concerts to track number of attendants.

It is good to confirm people are actually able to access tooling. That is all I need.