FAQ

FAQ

Short answers to the most common questions. For anything not here, see Troubleshooting.

What is CHUB?

A self-hosted web app that keeps a Plex library tidy. Point it at Radarr, Sonarr, Lidarr, and Plex and it handles the boring chores: renaming posters, finding duplicates, re-applying borders, searching for quality upgrades, cleaning up orphaned files, and more. See Home.

How is it different from DAPS?

Same module set, different direction. CHUB is a fork of DAPS with a refreshed UI, cooperative cancellation on 11 of the 12 modules, live updates via server-sent events, inline metadata editing with an audit trail, duplicate resolution, 6-hour system-health snapshots, token-bucket login rate limiter, SSRF guard, and expanded log redaction. See Credits for the full list.

Can I migrate my DAPS install to CHUB?

No. CHUB is a clean break — no data migration, no compatibility shims. Pull ghcr.io/chodeus/chub:latest into a fresh config directory and reconfigure. Your DAPS install keeps working on its own image alongside CHUB if you want to run both side-by-side during cut-over.

Is CHUB safe to expose on the public internet?

Not directly. CHUB is built for a private network — LAN or VPN. It has built-in auth (bcrypt + JWT), a login rate limiter, and SSRF protection, but no WAF, IP allow-list, or DDoS mitigation. For remote access, put it behind a reverse proxy with TLS plus a second auth layer (Authelia, Authentik, Cloudflare Access, etc.).

How do I reset the admin password?

Two options:

Run the reset command and restart CHUB:

docker compose run --rm chub python3 main.py --reset-auth

Or stop CHUB, delete the auth: block from config.yml, and start CHUB again. Either way the first-run form reappears.

Does CHUB work with a 4K-split Plex library?

Yes. Use nestarr to move items between split libraries based on ARR path mappings, and configure each Plex library separately under instances.plex. poster_renamerr also accepts a library_names list so you can scope each Plex instance.

Does CHUB support Lidarr?

Yes for upgradinatorr — full album search, artist grouping, and all three search modes (upgrade, missing, cutoff). There's no dedicated Lidarr UI because music library browsing is covered by Lidarr itself. Open an issue if you'd like a Lidarr-specific workflow built into CHUB.

How do I run a module right now instead of waiting for the schedule?

From the UI: Dashboard → New run, pick the module, Run.

From the API: POST /api/modules/{name}/execute.

How do I stop a running module?

From the UI: Settings → Jobs → click the running job → Cancel.

From the API: DELETE /api/modules/{name}/execution/{job_id}.

Cancellation is cooperative — the module has to check a cancel flag in its loop. 11 of the 12 modules do: poster_renamerr, poster_cleanarr, labelarr, jduparr, nohl, unmatched_assets, upgradinatorr, renameinatorr, health_checkarr, nestarr, sync_gdrive. border_replacerr doesn't yet — if you start a big run you have to restart the container to stop it.

Where does CHUB keep its data?

Everything under ${CONFIG_DIR} (default /config inside the container):

• config.yml — YAML config
• chub.db — SQLite database (users, jobs, media cache, edit history, health snapshots, poster cache)
• logs/ — per-module log files
• backups/ — zips from POST /api/backup

Your poster and media trees live on the volumes you mount separately.

How do I back it up?

Two options:

1. From the UI (or API): POST /api/backup returns a zip of config.yml + chub.db. It's also written to $CONFIG_DIR/backups/.
2. Filesystem: stop the container (for SQLite consistency), tar the config directory, restart.

How do I update?

docker compose pull
docker compose up -d

Schema changes are applied automatically on startup — no manual migrations.

Can I use CHUB with just one ARR instance?

Yes. Multi-instance support is optional — configure only the instances you actually run.

Does CHUB have an API key for integrations?

No dedicated API-key system. Use the JWT from POST /api/auth/login as a bearer token; bump auth.token_expiry_hours if you need long-lived automation. For inbound webhooks, use general.webhook_secret instead.

Do I need both Kometa and CHUB?

No. They solve different problems — Kometa manages Plex metadata and collections, CHUB manages poster file trees and media-asset chores. Many people run both. poster_renamerr is explicitly designed to consume Kometa's asset output.

Can I disable modules I don't use?

Yes. Just don't configure them and don't schedule them. Every module has a sidebar entry, but an unconfigured module simply won't do anything when triggered.

What's the container user / permissions model?

The image runs as dockeruser, with UID/GID set via PUID/PGID env vars. The defaults are 100 / 99 (matching Unraid). Most other Linux hosts want 1000 / 1000. Mount paths must be writable by whichever UID/GID you pick.

How do I contribute?

1. Open an issue first for anything non-trivial.
2. Fork, branch, PR against main.
3. Run linters (ruff for Python, eslint for JS) and build the frontend (npm run build) before pushing.
4. Keep PRs focused — one feature per PR.

Small fixes can skip the issue step.