kafka mysql数据源_streaming+kafka 数据来源mysql数据库

package spark

import java.util.Properties

import java.util.HashMap

import org.apache.kafka.clients.producer._

import org.apache.spark.SparkConf

import org.apache.spark.streaming._

import org.apache.spark.streaming.kafka._

import org.apache.spark.sql.SQLContext

import org.apache.spark.{ SparkContext, SparkConf }

import spark.bean.orders

object SelectFromOneTable {

def main(args: Array[String]) {

val Array(brokers, topic, wordsPerMessage) = Array("localhost:9092", "sun_test", "1")

val props = new HashMap[String, Object]()

props.put(ProducerConfig.BOOTSTRAP_SERVERS_CONFIG, brokers)

props.put(ProducerConfig.VALUE_SERIALIZER_CLASS_CONFIG,

"org.apache.kafka.common.serialization.StringSerializer")

props.put(ProducerConfig.KEY_SERIALIZER_CLASS_CONFIG,

"org.apache.kafka.common.serialization.StringSerializer")

props.put("serializer.class", "kafka.serializer.StringEncoder")

props.put("producer.type", "async")

val producer = new KafkaProducer[String, String](props)

val sparkConf = new SparkConf().setAppName("Spark SQL Test Case").setMaster("local")

val sparkContext = new SparkContext(sparkConf)

val sqlContext = new SQLContext(sparkContext)

val url = "jdbc:mysql://localhost:3306/sun_test?user=root&password=Sun@123";

val prop = new Properties();

val df = sqlContext.read.jdbc(url, "flow", prop).collect()

for (a

println(a)

val message = new ProducerRecord[String, String](topic, null, a.toString())

producer.send(message)

}

}

}