Kafka client for Scala

[chelios....@gmail.com]

Hi Guys,

Which is the recommend client in Scala for Kafka ?

I've found the following two 

https://github.com/cakesolutions/scala-kafka-client

https://github.com/akka/reactive-kafka

https://github.com/elodina/scala-kafka

I don't know which one to use ? Can someone please suggest!

Cheers

Chel

[Alan Reiner]

FYI, an indirect answer to your question is that I use the base java client code ("kafka-clients" maven module) from Scala.   Further, the lower-level API ("kafka_2.10" module, as of this writing) is actually in Scala already, though they are both straightforward to use in Scala:

This is what my pom.xml looks like (if you're using maven):

        <dependency>

            <groupId>org.apache.kafka</groupId>

            <artifactId>kafka-clients</artifactId>

            <version>0.9.0.1</version>

        </dependency>

        <dependency>

           <groupId>org.apache.kafka</groupId>

           <artifactId>kafka_2.10</artifactId>

           <version>0.9.0.1</version>

        </dependency>

Then I can access the libraries via:

// The following imports are for kafka tools in the "kafka-clients" maven dependency

import org.apache.kafka.clients.consumer.KafkaConsumer;

import org.apache.kafka.clients.producer.KafkaProducer;

import org.apache.kafka.clients.consumer.ConsumerRecords;

import org.apache.kafka.clients.consumer.ConsumerRecord;

import org.apache.kafka.clients.producer.ProducerRecord;

import org.apache.kafka.common.serialization.ByteArraySerializer

import org.apache.kafka.common.serialization.StringSerializer

import org.apache.kafka.common.TopicPartition

// The following imports are from kafka tools in the kafka_2.10 maven dependency

import kafka.consumer._

import kafka.api.{PartitionOffsetRequestInfo, OffsetRequest}

import kafka.common.TopicAndPartition

import kafka.client.ClientUtils

import kafka.utils.{ToolsUtils, CommandLineUtils}

These are the import that I happen to use in my project.  Note that the kafka_2.10 module imports are low-level tools that most people probably wouldn't need.  Stick with the high-level "kafka-clients" module.

I can't speak to the maturity of the "native" scala libraries, though I suspect most are just wrappers around the java library.