Updating

Updating Censorarr

Native Windows

Download the newer Censorarr-Setup-X.Y.Z.exe from GitHub Releases and run it over the existing installation.

Application files under C:\Program Files\Censorarr are replaced. Persistent data under C:\ProgramData\Censorarr is preserved.

Native Linux .deb

Download the newer package and install it over the current version:

sudo apt install ./Censorarr-X.Y.Z-linux-amd64.deb

The package replaces /opt/censorarr while preserving:

/var/lib/censorarr
/etc/default/censorarr

Then check:

sudo systemctl status censorarr
journalctl -u censorarr --since "5 minutes ago"

Normal source / Docker install

From the Censorarr project folder:

git pull
docker compose up -d --build

This updates source and recreates/rebuilds the container as needed.

Synology Container Manager

After replacing/pulling source:

1. Open the Censorarr project.
2. Rebuild/recreate the project when Dockerfile, entrypoint, compose, or dependency behavior changed.
3. A simple restart is sufficient only for changes that are already bind-mounted into the running container and do not affect the image/entrypoint.

When in doubt after an application version update, rebuild.

GPU Worker update

cd Censorarr/gpu-worker
git pull
docker compose down
docker compose up -d --build

If you use a separate clone on the GPU host, update that clone independently.

Preserve these directories

Do not delete persistent runtime data unless you intentionally want a reset.

Docker/source:

config/
work/
gpu-worker/models/

Native Linux:

/var/lib/censorarr

Native Windows:

C:\ProgramData\Censorarr

Before upgrading

Recommended:

• back up your persistent Censorarr configuration
• note any service/compose environment values
• confirm your worker token if using remote GPU
• keep a copy of custom profanity/exception files

After upgrading

For Docker/source, check:

docker compose logs --tail=100

and run:

docker exec -it censorarr python /app/selftest.py

For native Linux, use:

sudo systemctl status censorarr
journalctl -u censorarr --since "5 minutes ago"

If the GPU Worker was also updated, test:

curl -H 'X-Censorarr-Token: YOUR_TOKEN' http://GPU-IP:9000/health