Skip to content

[Quick Start] Building a simple recommendation engine with QBit (CallBack Blocking)

fadihub edited this page Jun 8, 2015 · 4 revisions

##overview

To really grasp QBit, one must grasp the concepts of a CallBack and queues. A CallBack is a way to get an async response in QBit from a microservice. You call a service method and it calls you back.

Building a simple recommendation engine with QBit - CallBack Blocking

This wiki will walk you through the process of building a simple recommendation engine with QBit, in this example things are going to be very simple, and you will notice that we are blocking on user load. This is bad for most apps, but we meant to show it so that we can explain how to fix it in the next example.

What you will build

You will build a simple recommendation engine with QBit; that will give a set of recommendations to users. When you run it you will get the following:

Recommendations for: Bob
	Recommendation{recommendation='Take a walk'}
	Recommendation{recommendation='Read a book'}
	Recommendation{recommendation='Love more, complain less'}
Recommendations for: Joe
	Recommendation{recommendation='Take a walk'}
	Recommendation{recommendation='Read a book'}
	Recommendation{recommendation='Love more, complain less'}
Recommendations for: Scott
	Recommendation{recommendation='Take a walk'}
	Recommendation{recommendation='Read a book'}
	Recommendation{recommendation='Love more, complain less'}
Recommendations for: William
	Recommendation{recommendation='Take a walk'}
	Recommendation{recommendation='Read a book'}
	Recommendation{recommendation='Love more, complain less'}

How to complete this guide

In order to complete this example successfully you will need the following installed on your machine:

Now that your machine is all ready let's get started:

https://github.com/fadihub/worker-callback-blocking.git

Once this is done you can test the service, let's first explain the process:

The process will be explained in more detail under [[Detailed Tutorial] Building a simple recommendation engine with QBit - CallBack Blocking. ] (https://github.com/advantageous/qbit/wiki/%5BDetailed-Tutorial%5D-Building-a-simple-recommendation-engine-with-QBit-(CallBack-Blocking))

User.java Listing

~/src/main/java/io.advantageous.qbit.example.recommendationengine/User

package io.advantageous.qbit.example.recommendationengine;

/* Domain object. */
public class User {

    private final String userName;


    public User(String userName){
        this.userName = userName;

    }

    public String getUserName() {
        return userName;
    }

}

RecommendationService.java Listing

~/src/main/java/io.advantageous.qbit.example.recommendationengine/RecommendationService

package io.advantageous.qbit.example.recommendationengine;




import io.advantageous.boon.Lists;
import io.advantageous.boon.cache.SimpleLRUCache;
import java.util.List;

public class RecommendationService {


    private final SimpleLRUCache<String, User> users =
            new SimpleLRUCache<>(10_000);

    public List<Recommendation> recommend(final String userName) {
        System.out.println("recommend called");
        User user = users.get(userName);
        if (user == null) {
            user = loadUser(userName);
        }
        return runRulesEngineAgainstUser(user);
    }

    private User loadUser(String userName) {
        return new User("bob"); //stubbed out... next example will use UserService
    }

    private List<Recommendation> runRulesEngineAgainstUser(final User user) {

        return Lists.list(new Recommendation("Take a walk"), new Recommendation("Read a book"),
                new Recommendation("Love more, complain less"));
    }


}

RecommendationServiceClient.java Listing

~/src/main/java/io.advantageous.qbit.example.recommendationengine/RecommendationServiceClient

package io.advantageous.qbit.example.recommendationengine;

import io.advantageous.qbit.reactive.Callback;

import java.util.List;

/**
 * @author  rhightower
 * on 2/20/15.
 */
public interface RecommendationServiceClient {


    void recommend(final Callback<List<Recommendation>> recommendationsCallback,
                   final String userName);
}

Recommendtion.java Listing

~/src/main/java/io.advantageous.qbit.example.recommendationengine/Recommendation

package io.advantageous.qbit.example.recommendationengine;


/* Domain object. */
public class Recommendation {

    private final String recommendation;

    public Recommendation(String recommendation) {
        this.recommendation = recommendation;
    }


    @Override
    public String toString() {
        return "Recommendation{" +
                "recommendation='" + recommendation + '\'' +
                '}';
    }
}

PrototypeMain.java Listing

~/src/main/java/io.advantageous.qbit.example.recommendationengine/PrototypeMain

package io.advantageous.qbit.example.recommendationengine;

import io.advantageous.boon.core.Sys;
import io.advantageous.qbit.service.ServiceQueue;

import java.util.List;

import static io.advantageous.boon.core.Lists.list;
import static io.advantageous.qbit.service.ServiceBuilder.serviceBuilder;
import static io.advantageous.qbit.service.ServiceProxyUtils.flushServiceProxy;
import static java.lang.System.out;

/**
 * Created by rhightower on 2/20/15.
 */
public class PrototypeMain {

    public static void main(String... args) {


        /* Create the service. */
        RecommendationService recommendationServiceImpl =
                new RecommendationService();

        /* Wrap the service in a queue. */
        ServiceQueue recommendationServiceQueue = serviceBuilder()
                .setServiceObject(recommendationServiceImpl)
                .build().startServiceQueue().startCallBackHandler();

        /* Create a proxy interface for the service. */
        RecommendationServiceClient recommendationServiceClient =
                recommendationServiceQueue.createProxy(RecommendationServiceClient.class);


        /* Call the service with the proxy. */
        recommendationServiceClient.recommend(out::println, "Rick");

        /* Flush the call. This can be automated, but is a core concept. */
        flushServiceProxy(recommendationServiceClient);
        Sys.sleep(1000);



        /* Lambdas gone wild. */

        List<String> userNames = list("Bob", "Joe", "Scott", "William");

        userNames.forEach( userName->
                        recommendationServiceClient.recommend(recommendations -> {
                            System.out.println("Recommendations for: " + userName);
                            recommendations.forEach(recommendation->
                                    System.out.println("\t" + recommendation));
                        }, userName)
        );



        flushServiceProxy(recommendationServiceClient);
        Sys.sleep(1000);

    }
}

build.gradle Listing

apply plugin: 'idea'
apply plugin: 'java'
apply plugin: 'application'


sourceCompatibility = 1.8
version = '1.0'

repositories {
    mavenLocal()
    mavenCentral()
}

sourceCompatibility = JavaVersion.VERSION_1_8
targetCompatibility = JavaVersion.VERSION_1_8


mainClassName = "io.advantageous.qbit.example.recommendationengine.PrototypeMain"

dependencies {
    compile group: 'io.advantageous.qbit', name: 'qbit-jetty', version: '0.7.2'
    compile group: 'javax.inject', name: 'javax.inject', version: '1'
    compile('org.springframework.boot:spring-boot-starter-web:1.2.1.RELEASE') {
        exclude module: 'spring-boot-starter-tomcat'
    }
    compile 'org.eclipse.jetty:jetty-webapp:9.+'
    compile 'org.eclipse.jetty:jetty-jsp:9.+'

    testCompile "junit:junit:4.11"
    testCompile "org.slf4j:slf4j-simple:[1.7,1.8)"
}

Test The Service

With your terminal cd worker-callback-blocking

then gradle clean build and finally gradle run you should get the following:

Recommendations for: Bob
	Recommendation{recommendation='Take a walk'}
	Recommendation{recommendation='Read a book'}
	Recommendation{recommendation='Love more, complain less'}
Recommendations for: Joe
	Recommendation{recommendation='Take a walk'}
	Recommendation{recommendation='Read a book'}
	Recommendation{recommendation='Love more, complain less'}
Recommendations for: Scott
	Recommendation{recommendation='Take a walk'}
	Recommendation{recommendation='Read a book'}
	Recommendation{recommendation='Love more, complain less'}
Recommendations for: William
	Recommendation{recommendation='Take a walk'}
	Recommendation{recommendation='Read a book'}
	Recommendation{recommendation='Love more, complain less'}

Summary

You have just built and tested a simple recommendation engine with QBit, see you in the next tutorial!

Tutorials

__

Docs

Getting Started

Basics

Concepts

REST

Callbacks and Reactor

Event Bus

Advanced

Integration

QBit case studies

QBit 2 Roadmap

-- Related Projects

Kafka training, Kafka consulting, Cassandra training, Cassandra consulting, Spark training, Spark consulting

Clone this wiki locally