Installation and Building

The Meta Merge service is distributed as a Docker container, which simplifies its installation and usage. You can pull the image from the GitHub Container Registry.

Savant-RS Version

The Meta Merge service is part of the Savant-RS project. The Savant-rs version used in Savant must match the version of the Meta Merge service.

How to check the savant-rs version used in Savant

You can check the Savant-rs version used in Savant by running the following command:

SAVANT_VERSION=0.5.10
docker run -it ghcr.io/insight-platform/savant-adapters-gstreamer:$SAVANT_VERSION


============
== Savant ==
============

Savant Version 0.5.10
Savant RS Version 1.4.0
DeepStream Version 7.0

So, you can see that the version of savant-rs is 1.4.0. Thus, you need to use the same version of the Meta Merge service.

We support the following tags for the Meta Merge service:

• latest - the latest version (do not use it, we use it for development purposes);

• ${SAVANT_RS}-rolling - the latest version of the Meta Merge service for the specific version of Savant-rs, e.g. 1.4.0-rolling;

• $SAVANT_VERSION - the specific version matching the version of Savant, e.g. 0.5.10;

• savant-latest - the latest version of the Meta Merge service for the latest version of Savant.

So, if you want to use the latest Savant, you probably need to use the ${SAVANT_RS}-rolling tag for the Meta Merge service. If you want to use the release version of Savant, you can use the $SAVANT_VERSION tag.

Image naming

• X86-64: savant-meta-merge-x86;

• ARM64: savant-meta-merge-arm64.

Using The Docker Container

Pull the Docker image:

docker pull ghcr.io/insight-platform/savant-meta-merge-x86:savant-latest

# or for arm64

docker pull ghcr.io/insight-platform/savant-meta-merge-arm64:savant-latest

Run the container with your configuration file:

docker run \
   -v /path/to/your/config.json:/opt/etc/configuration.json \
   -v /path/to/your/python/handlers:/opt/python \
   ghcr.io/insight-platform/savant-meta-merge-x86:savant-latest

# or for arm64

docker run \
   -v /path/to/your/config.json:/opt/etc/configuration.json \
   -v /path/to/your/python/handlers:/opt/python \
   ghcr.io/insight-platform/savant-meta-merge-arm64:savant-latest

Here, you need to:

1. Mount your configuration file into the container at /opt/etc/configuration.json

2. Mount your Python handler modules directory to make them accessible to the service at /opt/python

3. Ensure your configuration file references the correct paths for Python modules

The configuration file defines the ingress sources, egress destination, merge callbacks, and queue settings.

Note

Sometimes, you may need extra Python modules to be installed in the container. In this situation, you use the image as a base image and install the modules in the container in the Dockerfile.

Environment Variables

The Meta Merge service supports environment variable substitution in configuration files. You can use the following pattern:

docker run \
   -e PYTHON_MODULE_ROOT=/opt/python \
   -e HOME=/root \
   -e USER=root \
   -v /path/to/your/config.json:/opt/etc/configuration.json \
   -v /path/to/your/python/handlers:/opt/python \
   ghcr.io/insight-platform/savant-meta-merge-x86:savant-latest

Building From Source

If you prefer to build the service from source, you can do so using Docker or Cargo:

1. Clone the Savant RS repository:

   git clone https://github.com/insight-platform/savant-rs.git
   

2. Build the service with Docker:

   docker build -t meta_merge -f docker/services/Dockerfile.meta_merge .
   

3. Build the service with Cargo:

   cargo build --release -p meta_merge
   

   Note

   Because of various reasons we use dynamic linking in savant-rs. Thus, you need to copy the dependencies to the distribution. Consult with our Docker-based build to find out how to do it.

Development Setup

For development purposes, you can run the Meta Merge service directly from the source:

LOGLEVEL=info cargo run -p meta_merge services/meta_merge/assets/configuration.json

This approach is useful when developing custom Python handlers or modifying the service configuration during development.

Note

Make sure you have Python development dependencies installed and the correct Python version (3.8+) available in your development environment.