Skip to content

Prowlarr (Core Service)

Prowlarr is an indexer manager for the Arr ecosystem. It centralizes indexer configuration and syncs them to Sonarr, Radarr, Lidarr, and Whisparr.

Service Relationships

Classification Role
Core Service Indexer manager
Depends On None
Optional Sonarr, Radarr, Lidarr, Whisparr
Exposes UI Yes (Web UI)

Configuration in dumb_config.json

"prowlarr": {
  "instances": {
    "Default": {
      "enabled": false,
      "postgres_enabled": false,
      "postgres_main_db": "",
      "postgres_log_db": "",
      "process_name": "Prowlarr",
      "repo_owner": "Prowlarr",
      "repo_name": "Prowlarr",
      "release_version_enabled": false,
      "release_version": "latest",
      "clear_on_update": false,
      "exclude_dirs": [],
      "platforms": [],
      "suppress_logging": false,
      "auto_update": false,
      "auto_update_interval": 24,
      "auto_update_start_time": "04:00",
      "pinned_version": "",
      "port": 9696,
      "config_dir": "/prowlarr/default",
      "config_file": "/prowlarr/default/config.xml",
      "log_file": "/prowlarr/default/logs/prowlarr.txt",
      "command": [],
      "env": {}
    }
  }
}

Key Configuration Fields

  • postgres_enabled: Opt this instance into DUMB-managed PostgreSQL config. SQLite is the default; set this to true only when you want Prowlarr to use PostgreSQL.
  • postgres_main_db / postgres_log_db: Optional database-name overrides. When blank, DUMB uses prowlarr-main and prowlarr-log for the default instance, or unique instance-scoped names for additional instances.
  • port: Web UI port (default 9696).
  • pinned_version: Optional version pin for Prowlarr updates.
  • repo_owner / repo_name: GitHub repo used for releases or branch builds.
  • release_version_enabled / release_version: Use GitHub releases (e.g., latest, nightly, prerelease, or a tag).
  • clear_on_update: Clear the install directory before updating.
  • exclude_dirs: Directories to preserve when clearing.
  • platforms: Build platforms (auto‑defaults to ["dotnet"] when using branches).
  • config_dir, config_file, log_file: Paths for config and logs.

PostgreSQL database mode

When postgres_enabled is true, DUMB:

  • enables the bundled PostgreSQL service if needed;
  • creates the Prowlarr main/log databases in postgres.databases;
  • starts PostgreSQL before Prowlarr; and
  • writes the required PostgresUser, PostgresPassword, PostgresHost, PostgresPort, PostgresMainDb, and PostgresLogDb entries to Prowlarr's config.xml.

During onboarding, enabling postgres_enabled for Prowlarr is enough; you do not need to separately select PostgreSQL as an optional service.

This does not migrate existing SQLite data

Setting postgres_enabled: true changes the database backend Prowlarr starts with. It does not copy prowlarr.db into PostgreSQL.

If you enable this on an existing SQLite-backed Prowlarr instance without doing a manual migration, Prowlarr can start against fresh PostgreSQL databases and appear empty or newly initialized.

For an existing instance, use Database Migration on the service page. DUMB requires a rehearsal, initializes an isolated Prowlarr-owned PostgreSQL schema, imports and validates every table, and preserves SQLite for rollback. Prowlarr's upstream guide still classifies migration as unsupported. See SQLite to PostgreSQL Migration.

PostgreSQL is not a temporary toggle

There is no known supported migration path from PostgreSQL back to SQLite for Prowlarr. Treat postgres_enabled: true as a long-term database choice unless you are willing to recreate the Prowlarr instance from scratch.

DUMB can restore the preserved pre-cutover SQLite state, but it does not copy later PostgreSQL writes back into SQLite.

Repo-based installs and updates

Prowlarr now supports the same repo‑based controls as other DUMB services.

  • GitHub releases: Set release_version_enabled: true with a valid repo_owner/repo_name.
  • Default updater: Leave release_version_enabled: false to use the standard Arr updater.

GitHub sources take priority when enabled and are not a fallback. pinned_version only applies to the traditional Arr updater.

Integration with DUMB

  • Connect Prowlarr to Sonarr/Radarr/Lidarr/Whisparr via their API keys and base URLs.
  • Prowlarr can share indexers across all Arr instances to keep setups consistent.

AIOStreams search integration

AIOStreams can query a DUMB-managed Prowlarr instance for torrent and Usenet results:

  1. Copy Prowlarr's API key from Settings → General.
  2. In AIOStreams' built-in Prowlarr settings, enter http://127.0.0.1:9696, replacing 9696 with this instance's saved DUMB port.
  3. Enter the API key and choose the indexers AIOStreams may query.
  4. Add the built-in Prowlarr addon to the configuration on /stremio/configure.

This is an AIOStreams runtime integration, not a Prowlarr application entry that DUMB creates automatically. Prowlarr redirects NZB grabs rather than proxying them, so review indexer rules if searches and grabs can leave through different public IP addresses.

Automated Arr sync

DUMB can auto-configure Prowlarr applications for enabled Arr instances. It reads each Arr API key from the Arr config_file, waits for the services to come up, and creates the matching Prowlarr app entries (full sync). Apps are created using the Arr process_name.

Requirements:

  • Prowlarr instance is enabled.
  • Arr instances are enabled with a valid port and readable config_file.
  • Prowlarr API key is available in its config_file.

Core-service tags

DUMB tags Prowlarr application entries when the Arr instance has core_service set (single value or list):

Arr core service Tag applied in Prowlarr
decypharr decypharr
infinidysk infinidysk

These tags are created automatically if they do not exist.

Indexer routing via tags

Prowlarr uses tags to decide which indexers sync into which Arr apps. DUMB links each Arr app to the tag matching its core service(s), so Decypharr, InfiniDysk, and AltMount indexers stay separated.

When you add custom indexers in Prowlarr, make sure the correct tag is applied so they sync to the intended Arr apps.

Custom indexer sync (Decypharr)

When Decypharr is enabled, DUMB also syncs custom indexer definitions and creates indexers:

  • Zilean (Torznab) with a base URL that prefers the local Zilean instance when enabled, and defaults to https://zileanfortheweebs.midnightignite.me otherwise. You can override the URL in Prowlarr after it is created.
  • StremThru (Torznab) using the hosted endpoint

If Decypharr is enabled, DUMB automatically enables the public StremThru indexer.

For Whisparr applications, DUMB widens Prowlarr sync categories to include the Movie/TV roots used by the Zilean/StremThru custom definitions plus the common XXX categories. This lets the custom indexers sync to Whisparr while keeping the normal Decypharr/InfiniDysk/AltMount tag separation.

Custom definitions are pulled from the Prowlarr-Indexers repository and stored in <prowlarr_config_dir>/Definitions/Custom (or <prowlarr_config_dir>/indexer/Definitions/Custom when present).

Other custom indexers in the definition repository are available to add manually in Prowlarr.

Accessing the UI

  • Navigate to: http://<host>:9696

Installation troubleshooting

Before extracting an Arr archive, DUMB validates that tar can read the full gzip archive and logs both the downloaded archive size and free space in the installation filesystem. If validation or extraction fails, the structured DUMB error includes tar's stderr, its exit status, the archive size, and the available space. Messages such as unexpected end of file, No space left on device, Permission denied, or Read-only file system therefore remain visible even when /opt/prowlarr is ephemeral.

A Prowlarr preinstall failure no longer shuts down the DUMB API or Frontend. DUMB records the failure, completes the remaining preinstall work, and retries Prowlarr during normal service startup. If the retry also fails, Prowlarr stays stopped while the DUMB control plane remains available for logs and service management.

For repeated extraction failures, check the Docker host's storage and inode capacity at its Docker root directory. A full container removal/recreation deletes the failed archive under /opt, so preserve the complete DUMB log when requesting support.

If extraction reports Function not implemented, the archive and available space are not normally the cause. Ubuntu 26.04's security-hardened GNU tar uses the Linux openat2 interface to confine archive extraction. Older host kernels, container runtimes, or seccomp profiles can return ENOSYS for that interface even when the same installation worked with an older DUMB image. DUMB retries these failures with Python's security-filtered archive extractor, which rejects paths and links that escape /opt/prowlarr.

If both extractors fail, check the host's kernel, Docker, containerd, runc, and seccomp versions. Do not permanently disable seccomp as a workaround; updating the host runtime retains its security boundary and native GNU tar extraction.

Resources