SimpleHomelab/Deployrr v6.0.42 on GitHub

Released: 7 Nov 2026

GPU hardware acceleration now works on release builds, and WUD moves to the maintained getwud/wud image with an admin login.

Install

npm / npx (Recommended), needs Node.js and npm:

sudo npx @simplehomelab/deployrr@latest

curl standalone installer, no Node.js or npm needed:

sudo bash -c "$(curl -fsSL https://files.deployrr.app/install.sh)"

Then run deployrr.

Existing users: see Action required below before updating.

Highlights

GPU hardware acceleration (Plex, Jellyfin, Emby, Ollama and other GPU apps) now works on release builds. If you picked a GPU on v6.0.x, see Action required.

Action required

Back up first (Pro): Tools > Backups > Backup.

Reinstall GPU apps to get GPU acceleration

On v6.0.x, choosing a GPU during an app install printed "GPU acceleration configured", but no GPU devices were added to the app. The app ran on CPU. This affects every app installed with a GPU on v6.0.x: Chromium, digiKam, Dispatcharr, Double Commander, Emby, FileZilla, Jellyfin, Kasm, Lollypop, Ollama, Plex, Remmina and XPipe Webtop.

After updating:

Apps > Refresh Cache (or Prerequisites > Build Cache). This downloads the GPU files. Without it, the next GPU install shows an error and installs on CPU.
For each GPU app: Apps > Browse Apps > the app > Yes (Install) > Yes (Already Installed), then pick your GPU in GPU Selection. A stopped app can also be reinstalled from Apps > Manage Apps > the app > Reinstall.
Check the install output shows GPU acceleration configured and no red error.

Warning: Reinstalling recreates the container from the app's default compose file. Your app data (volumes) is kept, but manual edits to compose/<hostname>/<app>.yml are replaced. Copy any edits before reinstalling.

WUD (What's Up Docker): new image and admin login

Existing WUD installs keep the old, unmaintained fmartinou/whats-up-docker image until you reinstall WUD. Any licence.

Go to Apps > Browse Apps > wud (What's Up Docker) > Install > Yes (Install).
In the Already Installed dialog, select Yes.
Enter a WUD admin username (pre-filled with admin) and password, then confirm the password.
Open WUD and log in with that username and password.

WUD now has its own login. You need it even when WUD is behind Traefik with an auth provider.

Warning: No WUD settings are lost: the old setup stored no data. From now on WUD keeps its users and state in appdata/wud in your Docker folder.
WUD resets the admin username and password to the values stored by Deployrr every time it starts, so changes made in the WUD web UI don't last. To change them:

Pro: Tools > Secrets Editor, edit wud_admin_user or wud_admin_password, then restart WUD (sudo docker restart wud).
Any licence: reinstall WUD (steps 1 to 3 above) and enter the new values.

If you change the username, open Configuration > Users in WUD afterwards and delete the old user if it is still listed.

Fixed

GPU acceleration was not set up on release builds: the GPU compose snippets were not in the apps download, and Deployrr still printed "GPU acceleration configured". They now ship with the apps cache. If a GPU snippet is missing or can't be added, Deployrr shows an error naming the file and installs the app on CPU. (#88)
WUD (What's Up Docker) now uses the maintained getwud/wud image. The old fmartinou/whats-up-docker image is abandoned. New installs ask for a WUD admin username and password. Existing users: see Action required. (#80)

Known issues

NVIDIA transcoding on Jellyfin, Plex and Emby: choosing the NVIDIA GPU shows an error and installs the app on CPU, because the NVIDIA snippet clashes with the app's environment: settings. A fix will follow. Intel and AMD on these apps are not affected. NVIDIA on Ollama and Dispatcharr uses a different snippet without this clash, but hasn't been tested on NVIDIA hardware yet.

Need help? Ask on Discord: https://www.simplehomelab.com/discord