kafka同步触发流处理作业

py49o6xq  于 2021-06-06  发布在  Kafka
关注(0)|答案(1)|浏览(337)

我尝试一个简单的测试,我用Kafka连接和Spark
我编写了一个自定义的kafka connect来创建这个源记录

SourceRecord sr = new SourceRecord(null,
                    null,
                    destTopic,
                   Schema.STRING_SCHEMA,
                    cleanPath);

在Spark中我收到这样的信息

val kafkaConsumerParams = Map[String, String](
      "metadata.broker.list" -> prop.getProperty("kafka_host"),
      "zookeeper.connect" -> prop.getProperty("zookeeper_host"),
      "group.id" -> prop.getProperty("kafka_group_id"),
      "schema.registry.url" -> prop.getProperty("schema_registry_url"),
      "auto.offset.reset" -> prop.getProperty("auto_offset_reset")
    )
val messages = KafkaUtils.createDirectStream[String, String, StringDecoder, StringDecoder](ssc, kafkaConsumerParams, topicsSet)

val ds = messages.foreachRDD(rdd => {
          val toPrint = rdd.map(t => {
            val file_path = t._2

            val startTime = DateTime.now()

            Thread.sleep(1000 * 60)
            1
      }).sum()
        LogUtils.getLogger(classOf[DeviceManager]).info(" toPrint = " + toPrint +" (number of flows calculated)")
      })
    }

当我使用连接器将多条消息发送到所需的主题(在我的测试中,它有6个分区)时,sleep线程获取所有消息,但是同步地而不是异步地预执行它们。
当我创建一个简单的测试生成器时,休眠是异步完成的。
我还创建了两个简单的使用者,并尝试了连接器和生产者,这两个任务都是异步使用的,这意味着我的问题在于spark接收连接器发送的消息的方式。我不明白为什么当我从制作人那里发送任务时,它们的行为方式不同。
我甚至打印了spark收到的唱片,它们完全一样
制作人发送的记录

1: {partition=2, offset=11, value=something, key=null}
2: {partition=5, offset=9, value=something2, key=null}

连接发送的记录

1: {partition=3, offset=9, value=something, key=null}

我的项目中使用的版本是

<scala.version>2.11.7</scala.version>
    <confluent.version>4.0.0</confluent.version>
    <kafka.version>1.0.0</kafka.version>
    <java.version>1.8</java.version>
    <spark.version>2.0.0</spark.version>

依赖项

<dependency>
            <groupId>io.confluent</groupId>
            <artifactId>kafka-avro-serializer</artifactId>
            <version>${confluent.version}</version>
        </dependency>
        <dependency>
            <groupId>io.confluent</groupId>
            <artifactId>kafka-schema-registry-client</artifactId>
            <version>${confluent.version}</version>
        </dependency>
        <dependency>
            <groupId>org.apache.avro</groupId>
            <artifactId>avro</artifactId>
            <version>1.8.0</version>
        </dependency>
        <dependency>
            <groupId>org.apache.spark</groupId>
            <artifactId>spark-core_2.11</artifactId>
            <version>${spark.version}</version>
        </dependency>
        <dependency>
            <groupId>org.apache.spark</groupId>
            <artifactId>spark-sql_2.11</artifactId>
            <version>${spark.version}</version>
        </dependency>
        <dependency>
            <groupId>org.apache.spark</groupId>
            <artifactId>spark-streaming_2.11</artifactId>
            <version>${spark.version}</version>
        </dependency>
        <dependency>
            <groupId>org.apache.spark</groupId>
            <artifactId>spark-streaming-kafka_2.11</artifactId>
            <version>1.6.3</version>
        </dependency>
        <dependency>
            <groupId>org.apache.spark</groupId>
            <artifactId>spark-graphx_2.11</artifactId>
            <version>${spark.version}</version>
            <scope>provided</scope>
        </dependency>
        <dependency>
            <groupId>com.datastax.spark</groupId>
            <artifactId>spark-cassandra-connector_2.11</artifactId>
            <version>2.0.0-RC1</version>
        </dependency>
        <dependency>
            <groupId>org.scala-lang</groupId>
            <artifactId>scala-library</artifactId>
            <version>2.8.0</version>
        </dependency>
<dependency>
            <groupId>io.confluent</groupId>
            <artifactId>kafka-avro-serializer</artifactId>
            <version>${confluent.version}</version>
            <scope>${global.scope}</scope>
        </dependency>
        <dependency>
            <groupId>io.confluent</groupId>
            <artifactId>kafka-connect-avro-converter</artifactId>
            <version>${confluent.version}</version>
            <scope>${global.scope}</scope>
        </dependency>
        <dependency>
            <groupId>org.apache.kafka</groupId>
            <artifactId>connect-api</artifactId>
            <version>${kafka.version}</version>
        </dependency>
chhkpiq4

chhkpiq41#

我们不能异步运行spark kafka流作业。但我们可以像Kafka消费者那样并行运行它们。为此,我们需要在中设置以下配置 SparkConf() :

sparkConf.set("spark.streaming.concurrentJobs","4")

默认情况下,其值为 "1" . 但我们可以把它改写成一个更高的值。
我希望这有帮助!

相关问题