subway-fare-calculator

Subway Fare Calculator

This example shows a subway fare calculator written in Java with the Restate SDK.

When a passenger takes the subway, he badges in at the start station and out at the end station. Based on that, the system calculates the trip fare.

The system also keeps track of how much the passenger has spent in total on the ongoing day.
Whenever the amount exceeds that of a day ticket, the system will charge a day ticket instead.

What does the example show?

• Join events from different Kafka topics: badge in and badge out.
• Stateful actors that get invoked by Kafka events, execute lightweight durable functions and interact with Restate's KV store. We track the total amount spent by a passenger and the current trip.
• Durable delayed actions: schedule tasks for later on and Restate makes sure they execute. Here we close trips at the end of the day.
• Durable side effects: execute and retry payments. Persist successfully executed payments.
• Slow retry loops (of minutes) don't block processing for events of other keys.
• This can run anywhere. Also on a serverless/FaaS platform like AWS Lambda.

Have a look at the CardTracker service to see how the system is implemented.

Download the example

• Via the CLI:

  restate example java-subway-fare-calculator && cd java-subway-fare-calculator

• Via git clone:

  git clone git@github.com:restatedev/examples.git
  cd examples/java/end-to-end-applications/subway-fare-calculator

• Via wget:

  wget /~https://github.com/restatedev/examples/releases/latest/download/java-subway-fare-calculator.zip && unzip java-subway-fare-calculator.zip -d java-subway-fare-calculator && rm java-subway-fare-calculator.zip

Running the example

1. Start the Kafka broker via Docker Compose: docker compose up -d.
2. Start the Restate Server with the Kafka broker configuration in a separate shell: restate-server --config-file restate.toml
3. Start the service: ./gradlew run
4. Register the service: restate -y deployments register localhost:9080
5. Let Restate subscribe to the Kafka topic badgein and badgeout and respectively the badgeIn and badgeOut handlers of the CardTracker service:

   curl localhost:9070/subscriptions -H 'content-type: application/json' -d '{
   "source": "kafka://my-cluster/badgein",
   "sink": "service://CardTracker/badgeIn",
   "options": {"auto.offset.reset": "latest"}
   }'
   
   curl localhost:9070/subscriptions -H 'content-type: application/json' -d '{
   "source": "kafka://my-cluster/badgeout",
   "sink": "service://CardTracker/badgeOut",
   "options": {"auto.offset.reset": "latest"}
   }'

Badge in by creating a Kafka producer:

docker exec -ti broker kafka-console-producer --bootstrap-server localhost:9092 --topic badgein --property parse.key=true --property key.separator=:

And send the message:

card-321:"Liverpool Street"

Later on badge out, by creating a Kafka producer in another terminal for the badgeout topic:

docker exec -ti broker kafka-console-producer --bootstrap-server localhost:9092 --topic badgeout --property parse.key=true --property key.separator=:

And send the message:

card-321:"Baker Street"

Play around by sending some more requests. You can check the current K/V state via the CLI:

restate kv get CardTracker card-321

As you badge in and out, the system will calculate the fare and the total amount spent by the passenger. Once the total amount spent exceeds the day ticket price, the system will charge a day ticket instead.