camel-lra-saga

This code example shows how to configure LRA SAGA SERVICE using Apache Camel

Apache Camel Documentation

https://camel.apache.org/components/4.4.x/eips/saga-eip.html#_using_the_lra_saga_service

Pre requisite

We need a running LRA Coordinator, to which our application can connect.

There is a lack of documentation from both apache and narayana on how to run LRA Coordinator, which is very dissapointing. But I figured it out through many searches.

To run LRA you need an execuatable JAR. One is provided by quarkus. Clone this repo:

https://github.com/jbosstm/lra-coordinator-quarkus

Run:

mvn clean package

You can run the application by running

java -jar target/quarkus-app/quarkus-run.jar

By default app runs on port 8080 .You will get something like this:

C:\Users\...\Desktop\Saga\LRA\lra-coordinator-quarkus-main>java -jar target/quarkus-app/quarkus-run.jar
__  ____  __  _____   ___  __ ____  ______
 --/ __ \/ / / / _ | / _ \/ //_/ / / / __/
 -/ /_/ / /_/ / __ |/ , _/ ,< / /_/ /\ \
--\___\_\____/_/ |_/_/|_/_/|_|\____/___/
2024-03-01 09:32:33,746 INFO  [io.quarkus] (main) lra-coordinator-quarkus 1.0.0-SNAPSHOT on JVM (powered by Quarkus 3.8.0) started in 1.085s. Listening on: http://0.0.0.0:8080
2024-03-01 09:32:33,752 INFO  [io.quarkus] (main) Profile prod activated.
2024-03-01 09:32:33,752 INFO  [io.quarkus] (main) Installed features: [cdi, rest-client-reactive, resteasy-reactive, resteasy-reactive-jackson, smallrye-context-propagation, smallrye-fault-tolerance, smallrye-openapi, swagger-ui, vertx]

Verify LRA Coordination status

If you are able to access this link, Everything is fine and you can proceed to coding.

http://localhost:8080/lra-coordinator

Code Walkthrough

In this example we will create an tracking service, using Spring Boot and Apache Camel DSL. We will create 1 route direct:createTracking with 2 saga routes for compensation and completion

But before that while creating route definition, we need to configure LRASagaService

        org.apache.camel.service.lra.LRASagaService sagaService = new org.apache.camel.service.lra.LRASagaService();
        sagaService.setCoordinatorUrl("http://localhost:8080");
        sagaService.setLocalParticipantUrl("http://0.0.0.0:8005");
        getContext().addService(sagaService);

Few things to keep in mind:

1. setCoordinatorUrl corresponds to host and port where LRA Coordinator is running.
2. setLocalParticipantUrl corresponds to how LRA Coordination looks at the host. As you can see below while starting, LRA is listnening on host 0.0.0.0 and our service is running on port 8005. Note: there is no context part for our endpoints, otherwise we need to include that too.

__  ____  __  _____   ___  __ ____  ______
 --/ __ \/ / / / _ | / _ \/ //_/ / / / __/
 -/ /_/ / /_/ / __ |/ , _/ ,< / /_/ /\ \
--\___\_\____/_/ |_/_/|_/_/|_|\____/___/
2024-03-01 09:32:33,746 INFO  [io.quarkus] (main) lra-coordinator-quarkus 1.0.0-SNAPSHOT on JVM (powered by Quarkus 3.8.0) started in 1.085s. Listening on: http://0.0.0.0:8080
2024-03-01 09:32:33,752 INFO  [io.quarkus] (main) Profile prod activated.

Route Definition

        from("direct:createTracking")
                        .saga().timeout(10, TimeUnit.SECONDS).to("direct:newTracking");

        from("direct:newTracking").saga()
                .propagation(SagaPropagation.MANDATORY)
                .timeout(10, TimeUnit.SECONDS)
                .option("id", header("id"))
                .compensation("direct:cancelTracking")
                .completion("direct:completeTracking")
                .transform().header(Exchange.SAGA_LONG_RUNNING_ACTION)
                .bean(createTrackingService, "createTracking")
                .end();

        from("direct:completeTracking")
                .transform().header(Exchange.SAGA_LONG_RUNNING_ACTION)
                .bean(createTrackingService, "completeTracking");

        from("direct:cancelTracking")
                .transform().header(Exchange.SAGA_LONG_RUNNING_ACTION)
                .bean(cancelTrackingService, "cancelTracking");

Participant Definition

    @PutMapping(value = "lra-participant/complete", produces = "application/json")
    public ResponseEntity<String> completeTracking(){
        try {
            System.out.println("Completing Tracking");
            return ResponseEntity.status(HttpStatus.OK).body(producerTemplate.requestBody("direct:completeTracking","").toString());
        } catch (Exception e) {
            return ResponseEntity.status(HttpStatus.BAD_REQUEST).body(null);
        }
    }
    @PutMapping(value = "lra-participant/compensate", produces = "application/json")
    public ResponseEntity<String> compensateTracking(){
        try {
            System.out.println("compensate Tracking");
            return ResponseEntity.status(HttpStatus.OK).body(producerTemplate.requestBody("direct:cancelTracking","").toString());
        } catch (Exception e) {
            return ResponseEntity.status(HttpStatus.BAD_REQUEST).body(null);
        }
    }

Logic of execution remains same as in memory saga service, on fail compensate is called, while on completion complete is called.

How to Run this Application

This is a simple maven application, just clone the repo and open cmd and run:

To Compile:

mvn clean install

To run the spring application

mvn clean compile exec:java

How to Test the Application

Open postman and hit this GET endpoint:

localhost:8006/createTracking

Note: By Default application run on 8005 port. Which can be changed under application.yml file

server:
  port: 8005

You can also check the running LRA client status in this URL

http://localhost:8080/lra-coordinator