Cloudron Forum

@adhodgson said: I'd love to see something like Mailman or Sympa integrated into Cloudron.

Hello Cloudron team, Since the recent Cloudron update, most of my applications no longer appear in Profile → App Passwords → Add app password → App. I have more than 20 sites, but the dropdown only shows five applications, plus Email (IMAP/SMTP). Previously, my applications appeared in this list. One affected application is Mailefy: Cloudron version: 10.1.5 App: WordPress (Developer) 7.1.3 Package: org.wordpress.unmanaged.cloudronapp@3.16.4 The application's SFTP dialog displays the server, port 222, username, and a “Use an app password” link. However, following that link opens the password creation form without Mailefy in the dropdown, so I cannot create its SFTP password. Steps to reproduce: Open the affected application's SFTP dialog. Click “Use an app password”. Open the App dropdown. The application is missing. Expected behavior: Applications that support SFTP should be available when creating an app password. I found an older report about missing applications in this dropdown, but I do not know whether it is related: https://forum.cloudron.io/topic/4231/problem-with-app-passwords Is this a known issue in 10.1.5? Is there a supported workaround, such as creating the app password through the CLI or API? I can provide screenshots showing the application details, SFTP dialog, and incomplete dropdown. Thank you.

On Cloudron 10.1.5, clicking the stop button on a running integrity check gives: Unhandled error. Status: 500. unkown: Expected values to be strictly equal: 'number' !== 'string' with onStopIntegrityCheck() in the trace. The check keeps running. It looks like the task ID is sent as a number where a string is expected. Workaround: on the server, systemctl stop box-task-<task id> stops it cleanly. The task ID is shown by ps aux | grep taskworker. Note also: "unkown" in the error message instead of "unknown".

@robi said: @marcusquinn said: On 26.04, follow the guide's MySQL 8.4 step (stop box, install mysql-server, re-run the box setup script, start box). After that, cloudron-support --troubleshoot passed cleanly and every app came back with the same state Can you expand to include the path/to/the/box/setup-script? https://docs.cloudron.io/guides/upgrade-ubuntu-26#post-upgrade

No, it was not whitelisted, and I've never came across any issue regarding the IP. Changing the "from" address in the app's settings from noreply@example.com to support@example.com did the trick. Don't know whether something with Cloudron (maybe with the last major update to the built-in mail server) caused the trouble or if I somehow misconfigured the app by accident. Or if any of the updates to Ghost itself caused the change...

I've packaged Headscale, the open-source, self-hosted Tailscale control server, as a community app. Your devices keep using the official Tailscale apps (Linux, macOS, Windows, iOS, Android), but join a private WireGuard mesh coordinated by your Cloudron instead of Tailscale's hosted service. [image: Headscale on Cloudron] Install: add this catalog under Settings → Community apps, then install Headscale from the App Store: https://raw.githubusercontent.com/marcusquinn/cloudron-headscale-app/main/CloudronVersions.json or cloudron install --versions-url <that URL> --location headscale.example.com. What fits in with Cloudron Single sign-on: tailscale up --login-server https://headscale.example.com opens a Cloudron login. The app's access control decides who can add devices, and Headscale creates a user for each person on first sign-in. Pre-auth keys for servers and headless devices, created with the headscale CLI in the app's Terminal. One HTTPS domain carries the control protocol, Noise and the embedded DERP relay through Cloudron's reverse proxy. The only extra port is STUN on UDP 3480 by default (3478 is Cloudron's TURN port, and the NetBird package's optional STUN port defaults to 3479). Backups: the SQLite database, server keys and config.yaml live in /app/data, so a restore brings back the same server identity and enrolled devices keep working. Good to know Headscale has no web interface. You manage users, devices, routes and exit nodes with headscale in the Terminal; the README has the common commands. A sign-in link only works while tailscale up is still waiting, and Headscale replaces it every 15 minutes. If you see "Gone: Your session has expired", run tailscale up again. Devices connect directly where they can and fall back to the embedded relay. Tailscale's public relays stay in the config as a second fallback; you can remove them to rely only on your server. Disabling the STUN port disables the embedded relay. config.yaml is yours to edit (DNS, ACL policy, relays). The package reapplies only the keys it owns: server URL, listeners, the relay switch and the SSO block. The image uses the checksum-verified upstream release binary on a digest-pinned Cloudron base image. Images and the catalog carry GitHub build provenance. Tested on Cloudron 9.1 with Tailscale 1.104.1 clients: enrolment by SSO and pre-auth key, direct connections, relay-only connections through the embedded DERP server, restart, backup and restore, and an update between builds. Links Package, setup guide and issues: https://github.com/marcusquinn/cloudron-headscale-app Headscale docs: https://headscale.net/stable/ This is a first release, so feedback is welcome, especially from people running it with many devices, subnet routers or exit nodes. I maintain the package independently of the Headscale project and I'm not affiliated with Tailscale Inc.; please report packaging problems on the package repo rather than upstream.

@fbartels I packaged the headless server: https://forum.cloudron.io/topic/16152/headscale-self-hosted-tailscale-control-server-with-cloudron-sso-and-a-built-in-derp-relay You can ask your AI for any explainers, help with setup (instructional or directly with ssh), and diagnosing any issues. I don't particularly use it, but should all work, though — and use Netbird if you want a GUI and like their UX.

Thanks, I have removed it.

I moved most of my apps to on-premise servers and kept a handful on a Hetzner VPS to keep costs down. I'm using SFF computers like the HP Elitedesk 800 G3 Mini (under $100 on eBay nowadays) and it's a solid machine for hosting. Older ram like DDR3/DDR4 are easier to find/buy too.