Running ErsatzTV Legacy in Docker and Docker Compose

ErsatzTV Legacy is distributed as a Docker image hosted on the GitHub Container Registry at ghcr.io/ersatztv/ersatztv. Running ErsatzTV Legacy in Docker is the recommended approach for most users — it bundles the correct FFmpeg build, Python runtime, and all required system libraries into a self-contained image that can be managed with Docker Compose.

Image

The published image is ghcr.io/ersatztv/ersatztv. You can pin to a specific release tag (e.g. ghcr.io/ersatztv/ersatztv:26.6.0) or use latest to always pull the most recent release. The image is built on Ubuntu Noble and ships with a custom FFmpeg 7.1.1 build maintained by the ErsatzTV project (ghcr.io/ersatztv/ersatztv-ffmpeg:7.1.1). Separate Dockerfiles exist for linux/amd64 (default), linux/arm64, and linux/arm/v7, so the image is multi-architecture and Docker will automatically select the correct variant for your host when pulling from the registry.

Port Mapping

ErsatzTV Legacy listens on port 8409 by default for both the web UI and the IPTV/streaming endpoints. Map this port to the same port on your host so that your media clients can reach the IPTV stream and EPG URLs directly:

8409:8409

The UI and streaming port can be changed independently using environment variables (see below), but the standard single-port deployment is recommended unless you have a specific need to separate them.

Volume Mounts

ErsatzTV Legacy uses two container paths that you must account for:

Path	Purpose
`/config`	Persistent data — SQLite database (`ersatztv.sqlite3`), log files, secrets, artwork cache, search index, templates, and all generated configuration
`/transcode`	Scratch space for live transcoding segments — should be a tmpfs mount for best performance

Your media library directories must also be bind-mounted into the container. Mount them read-only wherever possible to prevent accidental modification.

Mounting /transcode as tmpfs keeps transcoding scratch files in RAM. This dramatically reduces disk I/O during live streaming and avoids wear on SSDs. If you omit the tmpfs declaration, Docker will write segment files to a regular container layer, which is slower and accumulates data on restarts.

Environment Variables

Variable	Default	Description
`TZ`	System default	Host timezone (e.g. `America/Chicago`, `Europe/London`). Controls the timezone used for schedule times and EPG generation.
`ASPNETCORE_ENVIRONMENT`	`Production`	Set to `Development` to enable additional ASP.NET Core developer diagnostics. Leave as `Production` in normal use.
`ETV_CONFIG_FOLDER`	Platform data dir	Override the path where ErsatzTV stores its config and database. Set automatically to `/config` inside Docker.
`ETV_TRANSCODE_FOLDER`	Platform temp dir	Override the path used for transcoding scratch files. Set automatically to `/transcode` inside Docker.
`ETV_INSTANCE_ID`	(unset)	A short identifier appended to channel IDs in XMLTV output. Set this when running multiple ErsatzTV instances so their EPG channel IDs do not collide in an aggregator like Channels DVR.
`ETV_UI_PORT`	`8409`	Override the port the web UI listens on.
`ETV_STREAMING_PORT`	`8409`	Override the port the IPTV streaming endpoint listens on. Defaults to the same value as `ETV_UI_PORT`.
`ETV_BASE_URL`	(unset)	Set a URL base path when running behind a reverse proxy at a sub-path (e.g. `/ersatztv`).
`ETV_ALLOW_SHARED_PLEX_SERVERS`	(unset)	Set to any non-empty value to allow connecting to shared (non-owned) Plex servers.
`ETV_DISABLE_VULKAN`	`1` (in image)	Disables Vulkan in the NVIDIA pipeline. Enabled by default in the Docker image.
`ETV_MAXIMUM_UPLOAD_MB`	`10`	Maximum file upload size in megabytes. Increase if you need to upload larger files through the web UI.
`ETV_SLOW_DB_MS`	(unset)	Log a warning whenever a database query exceeds this threshold in milliseconds. Useful for diagnosing slow queries. Leave unset to disable.
`ETV_SLOW_API_MS`	(unset)	Log a warning whenever an API request exceeds this threshold in milliseconds. Leave unset to disable.
`ETV_JF_PAGE_SIZE`	`10`	Number of items fetched per page when scanning Jellyfin libraries. Reduce on low-memory devices if scanning causes memory pressure.
`ETV_JF_ENABLE_STATS`	(unset)	Set to any non-empty value to enable Jellyfin playback statistics reporting.
`ETV_NEXT_FOLDER`	(unset)	Override the folder path used by the ErsatzTV Next streaming engine.
`TRAKT__CLIENTID`	(unset)	Your personal Trakt API application Client ID. Required if you want to sync Trakt lists. Create an app at trakt.tv/oauth/applications.
`NVIDIA_VISIBLE_DEVICES`	(unset)	Required for NVIDIA GPU passthrough — set to `all` or a specific GPU index.
`NVIDIA_DRIVER_CAPABILITIES`	(unset)	Required for NVIDIA GPU passthrough — set to `compute,utility,video`.

Docker Compose Files

The repository ships three Compose files that can be layered together: a base file for the standard configuration and two overlay files for GPU acceleration variants.

Standard
NVIDIA GPU
VAAPI (Intel/AMD)

The base Compose file runs ErsatzTV Legacy with software transcoding. This is suitable for most installations and requires no special host configuration.

docker-compose.yml

services:
    ersatztv:
        image: ghcr.io/ersatztv/ersatztv:latest
        environment:
            TZ: America/Chicago
        ports:
            - "8409:8409"
        volumes:
            - ersatztv:/config
            #- /media/shared:/media/shared:ro
        tmpfs:
            - /transcode
volumes:
    ersatztv:

Uncomment and adjust the media volume mount (/media/shared:/media/shared:ro) for each directory you want to make available to ErsatzTV. Mount media read-only (:ro) unless you have a reason to write to those paths from within the container.

The NVIDIA overlay adds GPU device reservation and the required environment variables for NVIDIA hardware-accelerated transcoding. It is intended to be used together with the base Compose file using docker compose -f docker-compose.yml -f docker-compose.nvidia.yml up.

docker-compose.nvidia.yml

services:
    ersatztv:
        privileged: true
        environment:
            NVIDIA_VISIBLE_DEVICES: all
            NVIDIA_DRIVER_CAPABILITIES: compute,utility,video
        deploy:
            resources:
                reservations:
                    devices:
                        - driver: nvidia
                          count: 1
                          capabilities: [gpu]

The NVIDIA Container Toolkit must be installed on the host before this will work. Once the container is running, verify GPU detection under Settings → Troubleshooting → NVIDIA in the ErsatzTV UI.

The VAAPI overlay passes the DRM render node into the container for Intel Quick Sync Video and AMD VAAPI hardware-accelerated transcoding. It is intended to be used together with the base Compose file using docker compose -f docker-compose.yml -f docker-compose.vaapi.yml up.

docker-compose.vaapi.yml

services:
    ersatztv:
        devices:
            - /dev/dri/renderD128:/dev/dri/renderD128

If your system has multiple render nodes (e.g. both an iGPU and a dGPU), identify the correct one with ls -la /dev/dri/ on the host and adjust the device path accordingly. Verify VAAPI detection under Settings → Troubleshooting → VAAPI in the ErsatzTV UI.

Quick-Start docker run Examples

If you prefer docker run over Compose, these commands cover the most common configurations:

docker run -d \
  --name ersatztv \
  -e TZ=America/Chicago \
  -p 8409:8409 \
  -v ersatztv:/config \
  --tmpfs /transcode \
  -v /media/shared:/media/shared:ro \
  ghcr.io/ersatztv/ersatztv:latest

ARM Support

ErsatzTV Legacy provides Dockerfiles for two additional ARM architectures:

ARM64 (linux/arm64) — suitable for Raspberry Pi 4/5, Apple Silicon (via emulation), and other 64-bit ARM single-board computers. Uses mcr.microsoft.com/dotnet/aspnet:10.0-noble-arm64v8 as the .NET runtime base.
ARM32v7 (linux/arm/v7) — suitable for older Raspberry Pi 3 models and other 32-bit ARMv7 devices. Uses mcr.microsoft.com/dotnet/aspnet:10.0-noble-arm32v7 as the .NET runtime base.

Hardware transcoding (NVIDIA/VAAPI) is not available on ARM. Software transcoding works on both ARM variants, though performance will depend heavily on the device’s CPU speed.

Upgrading

To upgrade to a newer image, pull the updated image and recreate the container. Docker Compose makes this straightforward:

docker compose pull
docker compose up -d

ErsatzTV applies any required database migrations automatically on startup. There is no separate migration step. Always back up your /config volume before upgrading.

Single-Instance Enforcement

ErsatzTV Legacy uses a named system mutex (Global\ErsatzTV.Singleton.74360cd8985c4d1fb6bc9e81887206fe) to prevent more than one instance from running concurrently. If a second instance is started while the first is already running, it will immediately exit with the message:

Another instance of ErsatztTV is already running.

This applies even if you start two separate containers sharing the same /config volume. Only one container should ever have the config volume mounted at a time. If you need multiple independent ErsatzTV instances (e.g. for separate channel lineups), give each instance its own dedicated config volume and set distinct ETV_INSTANCE_ID values so their EPG identifiers remain unique.

Get Started

Core Concepts

Configuration

Scheduling

Media Server Integration

Streaming & Clients

Operations

Running ErsatzTV Legacy in Docker and Docker Compose

Image

Port Mapping

Volume Mounts

Environment Variables

Docker Compose Files

Quick-Start docker run Examples

ARM Support

Upgrading

Single-Instance Enforcement

Build docs developers (and LLMs) love

Get Started

Core Concepts

Configuration

Scheduling

Media Server Integration

Streaming & Clients

Operations

Documentation Index

​Image

​Port Mapping

​Volume Mounts

​Environment Variables

​Docker Compose Files

​Quick-Start docker run Examples

​ARM Support

​Upgrading

​Single-Instance Enforcement

Build docs developers (and LLMs) love

Image

Port Mapping

Volume Mounts

Environment Variables

Docker Compose Files

Quick-Start docker run Examples

ARM Support

Upgrading

Single-Instance Enforcement