covid-service-rs

Collects, processes and aggregates live COVID-19, and historical time series data from Johns Hopkins University. The aggregated data is then used to populate a mongoDB database which supplies data to https://github.com/mukundbhudia/covid-api and subsequently https://github.com/mukundbhudia/covid-web.

This is a conversion of the (now archived) JavaScript https://github.com/mukundbhudia/covid-service project into a Rust project.

The target environment (for my personal use case) will be a Raspberry Pi 4 Model B running Ubuntu 20.04 (aarch64-unknown-linux-gnu). However any environment that supports the Rust toolchain should suffice. The service will run as a cron task at regular intervals for example every 20 minutes.

Note: only Linux & OS X environments with aarch64 & x86_64 architectures have been tested though.

Pre-requisites

Rust

• Install Rust here.
• Minimum Supported Rust Version (MSRV) is any 2018 edition.

For Linux

• Make sure essential build tools (including compilers) are installed: sudo apt install build-essential.
• OpenSSL will be needed sudo apt install libssl-dev.
• pkg-config will be needed sudo apt install pkg-config.

For OSX

• Xcode and Xcode command line tools: xcode-select --install.

All platforms

• mongoDB - see: https://docs.mongodb.com/manual/installation/ for install instructions for your platform.

Development

• Within the repo directory run cargo r "DB_URI" "DB_NAME" where DB_URI is your mongo instance and DB_NAME is your mongo database.
• For example cargo r "mongodb://localhost:27017/" "covid19" for connection to local mongoDb instance to the database: covid19.

Production

• To make a production build, within the repo directory run cargo b --release.
• To run the release build, within the project directory run target/release/covid-service-rs "DB_URI" "DB_NAME" with the same arguments as in development.

Testing

• To run all your tests, within the repo directory run cargo t. This will run all tests in the /tests directory.
• To run test a specific test such as /tests/my_test.rs, within the repo directory run cargo t --test my_test.
• To run ignored tests, run cargo t -- --ignored.

Installation

• If there is a release asset for your target platform, you can run a command like wget -q -O tmp.zip https://github.com/mukundbhudia/covid-service-rs/releases/latest/download/covid-service-rs_aarch64-unknown-linux-gnu.zip && unzip -o tmp.zip && rm tmp.zip && strip covid-service-rs && mv covid-service-rs /home/ubuntu/.cargo/bin/covid-service-rs to have the binary placed in the cargo bin folder.

• For a fresh build and install on your target environment, run cargo install --path . within the repo directory. Then covid-service-rs will be moved to your cargo bin directory and will be callable throughout your host environment as a standard binary.

• To install as a cron task, an example entry under crontab -e to run the service every 20 minutes would be */20 * * * * /home/my-user-name/.cargo/bin/covid-service-rs "mongodb://localhost:27017/" "covid19"

• Subsequent upgrades would then only require a git pull and cargo install --path . to be run again to build and move the binary to cargo bin.

Advanced builds (cross-compilation)

• Run rustc --print target-list to see a list of targets you can build for.
• You can run rustup target add <my-target> to add the specific target to your rustup toolchain. This means you can now compile and build for that target.
• Running rustup show will show toolchains already installed on your host environment.
• Then running cargo b --release --target=<my-target> would create the build for that environment.

For more information on cross-compilation see https://rust-lang.github.io/rustup/cross-compilation.html.

Target schema

The intended mongo database should comprise of two collections:

• casesByLocation which holds data on every country and/or state along with the time series data. The stuct for this collection can be found as CaseByLocation in /src/schema.rs.
• totals which holds global data including global time series data and . The stuct for this collection can be found as GlobalCaseByLocation in /src/schema.rs.

Resources & Thanks

• To Johns Hopkins CSSE for the hard work providing and collating the data.
• To Our World in Data for their hard work in providing rich, clear and accurate data available on COVID-19.
• To the wonderful Rust community for their excellent book.