Troubleshooting

Troubleshooting

Start with Admin Panel > Server logs or docker logs dockerupdates: most problems are explained there.

I changed .env but nothing changed

Docker copies .env into the container only when it is created. Restart, Force update and docker restart keep the old values. Recreate it: docker compose up -d, or docker rm -f dockerupdates and the same docker run again. Also check whether that setting was saved in the web UI: a value saved in Settings takes precedence over .env (look for the Set here badge).

I forgot my username or password

If you changed them in Settings > Account: set RESET_LOGIN_CONFIG=true, recreate the container, sign in with ADMIN_USER / ADMIN_PASSWORD from .env, then set it back to false and recreate again.

If you never changed them: they are ADMIN_USER / ADMIN_PASSWORD in your .env (docker exec dockerupdates printenv ADMIN_USER).

"Too many failed attempts"

Wait for the lockout window (15 minutes by default), or unlock it from another signed-in browser in Admin Panel > IP access. If everyone behind your reverse proxy is locked out together, set Trust a reverse proxy (TRUST_PROXY).

Everyone is signed out after every restart or update

No session secret is set. Set one in Settings > Server & access or SESSION_SECRET.

The header shows the wrong IP

The IP is detected automatically. A wrong one usually comes from an old HOST_IP in .env (shown in orange in Settings > Server & access) or from an IP typed in Settings. Click Auto, or remove HOST_IP from .env. See HOST_IP.

"Auth required" on a container

The image is private. Add credentials: see Private registries.

"Check failed" on a container

Hover the badge for the reason. Usual causes: the tag no longer exists, the registry is down or rate-limited (Docker Hub), or DockerUpdates has no internet access (DNS / firewall).

"Could not use Icon URL"

• opens a web page, not an image: use the direct link to the image file (right-click the image > Copy image address). Links to GitHub / GitLab file pages are converted automatically.
• did not answer within 15 s: check the URL and that DockerUpdates can reach the internet.

"Port … is already in use"

The host port you typed in Add / Edit is used by another container (running, or stopped with the port reserved) or by a service on the server. Pick another host port.

The console does not open

The console shows the reason in red inside the terminal:

• Could not open the console connection: the connection never reached DockerUpdates. Almost always the reverse proxy does not allow WebSockets: in Nginx Proxy Manager turn on Websockets Support in the proxy host; in Cloudflare turn on Network > WebSockets. See Reverse proxy.
• Origin not allowed: the proxy changes the Host header. Pass it unchanged, or add the public address to Allowed origins.
• Your session expired: sign in again.
• Session closed · Connection lost after about a minute without typing: the reverse proxy closes idle connections. Set Console keep-alive to 25 in Settings > Server & access (CONSOLE_WS_KEEPALIVE).
• The container is not running / has no shell: the container must be running and have sh or bash (minimal images such as distroless or scratch have none).

Refused connections are also written to Admin Panel > Server logs as console refused: …, with the origin and host the server received.

Origin not allowed / Missing Origin header

Your reverse proxy changes the Host header. Pass it unchanged, or add the public URL to ALLOWED_ORIGINS.

DockerUpdates does not start

Check docker logs dockerupdates:

• ADMIN_USER and ADMIN_PASSWORD must be set: add them to .env.
• ADMIN_PASSWORD is too weak: 8+ characters, not a default value, not the username.
• SESSION_SECRET must be at least 32 characters: use openssl rand -hex 32.
• connect ENOENT /var/run/docker.sock: mount the Docker socket (-v /var/run/docker.sock:/var/run/docker.sock).

Settings or history disappeared after an update

The data volume is not mounted: add -v ~/dockerupdates/data:/app/backend/data (or ./data:/app/backend/data in Compose).