qciconnect-miniserver 0.4.2

Mini server for local development

 

Miniserver

This python package is used to locally execute the compilers and simulators of the QCI Connect platform of the German Aerospace Center (DLR). It is a component of the full QCI Connect SDK and can interface as a server with the API Client, substituting the QCI Connect platform.

It allows to

1. locally develop algorithms with the goal to integrate them to the platform,
2. prototype quickly, and
3. run algorithms on the local computer using the standard interfaces of the QCI Connect API Client.

Quick Start (Pip Installation)

The miniserver requires Docker to be installed.

• Run pip install qciconnect-miniserver qciconnect-client
• Build compiler images that you want to use (see instructions below for building the example compilers included in the miniserver repository)
• Start the server with the compiler path miniserver -c path_to_compilers_folder
• Connect API Client with the server as described in the example notebook

Working on Source Code (Dev Mode)

Clone the repository (or download the .tar.gz file), initialize the git submodules, then install dependencies.

We recommend using ultraviolet (installation guide).

git submodule update --init
uv sync

You can now activate the .venv in the source directory or use uv to run commands in the .venv, e. g., uv run python my_script.py.

Using the Miniserver (Dev Mode)

The following steps provide a quick start guide for programmatically interacting with the local development environment.

Build Compiler Image

First, verify that Docker (including the Docker Compose plugin) is installed, then build the compiler step images as follows:

cd deployment/docker-dev
docker compose build --no-cache

Start Miniserver

Next, you’re ready to start the miniserver:

uv sync
source .venv/bin/activate
miniserver

or

uv run miniserver

Troubleshooting

Miniserver

In case you need to restart the miniserver just stop it through CTR+C or

miniserver stop

It could be that miniserver does not restart properly due to compiler-containers still running from previous sections

docker rm -v -f <compiler-container-list>