naludaq-rs 0.2.12

Rust backend for NaluDAQ

NaluDAQ_rs

Rust backend for NaluDaq.

Setup

Make sure the Rust toolchain is installed.

Clone the repository to a suitable location:

git clone https://github.com/NaluScientific/naludaq_rs.git

Done!

Building the backend for python

pip install maturin
maturin build

Building the Backend

The backend is easy to build:

cd naludaq_rs
cargo build --release

Developers may ommit the --release flag for unoptimized builds.

Building for RPI4

Building using docker:

docker build . -t cc/rpi4

to make the container, the build the app with:

docker run --rm -v ${pwd}:/app cc/rpi4

Running the Backend

To run the backend, run the following command:

cargo run  --release -- [<ROOT>] [--addr <ADDR>] [-d | --debug]  [--api]

Where the arguments are as follows:

• <ROOT> is the root directory to run the server in. By default, it is the current working directory of the terminal it was run from.
• --addr <ADDR> is the address to bind the server to in the format IP:PORT. If unspecified, the server is bound to an open port on the loopback address.
• -d | --debug shows additional debug messages.
• --api will open the Swagger UI in the system browser.

Documentation

The documentation can be built using the following command:

cargo doc --document-private-items --no-deps

The API

NaluDAQ_rs is controlled through its REST API over HTTP. This allows it to be controlled from any machine connected to a network.

The backend can be controlled on the fly via Swagger. Use the --api flag when starting the backend to display the Swagger UI.

Alternatively, the (non-interactive) Swagger UI may be viewed without the backend by running the following command:

cargo run --bin api -- [--output OUTPUT_FILE]

If the --output argument is provided, the OpenAPI JSON data describing the API is written to the file and the program will immediately exit.

Developers

Profiling

Use the build profile called `release-with-debug for building release mode with debug symbols.

cargo build --profile release-with-debug

On Windows machines with Intel processors you can use Intel VTune to profile the backend.

Building the Documentation

cargo doc --open

Tweaking Dependencies

The crate depends on a couple crates published to crates.io. If you are touching these dependencies and wish to test out your changes without publishing the crates, clone the dependencies and change these lines in Cargo.toml:

# naluacq = "X.X.X"
# ft60x_rs = "X.X.X"
naluacq = { "path" = "path/to/naluacq" }
ft60x_rs = { "path" = "path/to/ft60x-rs" }

Architecture

Below is a non-exhaustive diagram of the architecture of the backend highlighting only the main components and their interactions.

graph LR

    Board[Board]
    Connection[Connection]
    Disk[Disk]
    subgraph workers [Workers]
        Worker_Connection[Connection Worker]
        Worker_Package[Package Worker]
        Worker_Storage[Storage Worker]
        Worker_Answer[Answer Worker]
        Worker_Answer_Buffer[Answer Buffer]
    end
    subgraph server [Server Endpoints]
        Endpt_Configure_Workers[Configure Workers]
        Endpt_Send_Command[Send Command]
        Endpt_Recv_Answer[Receive Answer]
        Endpt_Fetch_Event[Fetch Event]
        Endpt_Manage_Connection[Manage Connection]
    end
    Board -->|Readout/Answer| Connection
    Connection -->|Board Output| Worker_Connection
    Connection -->|Command| Board
    Worker_Connection -->|Board Output| Worker_Package
    Worker_Package -->|Event| Worker_Storage
    Worker_Package -->|Answer| Worker_Answer
    Worker_Storage -->|Event| Disk
    Worker_Answer -->|Answer| Worker_Answer_Buffer

    Endpt_Send_Command -->|Command| Connection
    Worker_Answer_Buffer -->|Answer| Endpt_Recv_Answer
    Disk -->|Event| Endpt_Fetch_Event
    Endpt_Manage_Connection -->|Configuration| Connection
    Endpt_Configure_Workers -->|Configuration| Worker_Package
    Endpt_Configure_Workers -->|Configuration| Worker_Storage
    Endpt_Configure_Workers -->|Configuration| Worker_Answer