Quick start

Here is how to get started:

ezrules is composed of two parts: ezrule manager and ezrule executor. The former is responsible for the UI for rule management that analysts can use to modify and maintain rules. The latter is a standalone rule execution enviroment that handles processing of events(e.g. transactions).

The below tutorial shows how to get started locally and on a on-premise k8s cluster. For both cases you will need a postgres database running and availbale to the both manager and executor.

If you do not have a local postgres instance running, you may need docker running on your machine.

Local setup

Preparing the environment

Install ezrules with

pip install ezrules

First, get and get the database running:

docker run --rm -it --name postgresql16 -p 5432:5432 -e POSTGRES_PASSWORD=root -e POSTGRES_HOST_AUTH=trust -e LC_CTYPE=en_US.UTF-8 -e PG_HOST=localhost postgres:16.0-alpine3.18

This will pull postgres16 image and run it in a container.

Next, install the dependancies with pipenv install and activate the new enviroment with pipenv shell.

First, we need to setup the environment.

ezrules init-db --db-endpoint postgresql://postgres:root@localhost:5432/postgres

And then add a user to login with:

ezrules add-user --db-endpoint postgresql://postgres:root@localhost:5432/postgres --user-email [email protected] --password 123

Running the application

In separate ternminal windows, run the manager and an executor instance, but before that, export the following env variables:

export DB_ENDPOINT=postgresql://postgres:root@localhost:5432/postgres
export APP_SECRET=123

ezrules manager --db-endpoint postgresql://postgres:root@localhost:5432/postgres

and the executor

ezrules evaluator --db-endpoint postgresql://postgres:root@localhost:5432/postgres

By default, the manager will start at http://0.0.0.0:8888. Go there and login with the details above. Then click Add rule and create a test configuration as shown on the screenshot:

Click Submit.

The new rule will be automatically picked up by the evaluator running in the second terminal. Test your new rule with

curl -X POST -H "Content-Type: application/json" -d '{"age": 25, "amount": 950}' http://localhost:9999/evaluate

This will return ["CANCEL"]

K8s cluster