我正在尝试提交一个spark streaming + kafka作业,它只是从kafka主题中读取字符串行。但是,我收到以下异常
15/07/24 22:39:45 ERROR TaskSetManager:阶段2.0中的任务0失败了4次;堕胎 线程“Thread-49”中的异常org.apache.spark.SparkException:作业因阶段失败而中止:阶段2.0中的任务0失败4次,最近失败:阶段2.0中丢失任务0.3(TID 73,10.11.112.93) :java.lang.NoSuchMethodException:kafka.serializer.StringDecoder。(kafka.utils.VerifiableProperties) java.lang.Class.getConstructor0(Class.java:2892) java.lang.Class.getConstructor(Class.java:1723) org.apache.spark.streaming.kafka.KafkaReceiver.onStart(KafkaInputDStream.scala:106) org.apache.spark.streaming.receiver.ReceiverSupervisor.startReceiver(ReceiverSupervisor.scala:121) org.apache.spark.streaming.receiver.ReceiverSupervisor.start(ReceiverSupervisor.scala:106) org.apache.spark.streaming.scheduler.ReceiverTracker $ ReceiverLauncher $$ anonfun $ 9.apply(ReceiverTracker.scala:264) org.apache.spark.streaming.scheduler.ReceiverTracker $ ReceiverLauncher $$ anonfun $ 9.apply(ReceiverTracker.scala:257) org.apache.spark.SparkContext $$ anonfun $ runJob $ 4.适用(SparkContext.scala:1121) org.apache.spark.SparkContext $$ anonfun $ runJob $ 4.适用(SparkContext.scala:1121) org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:62) org.apache.spark.scheduler.Task.run(Task.scala:54) org.apache.spark.executor.Executor $ TaskRunner.run(Executor.scala:177) java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145) java.util.concurrent.ThreadPoolExecutor中的$ Worker.run(ThreadPoolExecutor.java:615) java.lang.Thread.run(Thread.java:745)
当我检查DSE使用的spark jar文件时,我发现它使用的kafka_2.10-0.8.0.jar确实有这个构造函数。不确定导致错误的原因。这是我的消费者代码
val sc = new SparkContext(sparkConf)
val streamingContext = new StreamingContext(sc, SLIDE_INTERVAL)
val topicMap = kafkaTopics.split(",").map((_, numThreads.toInt)).toMap
val accessLogsStream = KafkaUtils.createStream(streamingContext, zooKeeper, "AccessLogsKafkaAnalyzer", topicMap)
val accessLogs = accessLogsStream.map(_._2).map(log => ApacheAccessLog.parseLogLine(log).cache()
更新此异常似乎仅在我提交作业时才会发生。如果我通过粘贴代码来使用spark shell来运行作业,那么它可以正常工作
答案 0 :(得分:1)
我的自定义解码器遇到了同样的问题。我添加了以下构造函数,它解决了这个问题。
public YourDecoder(VerifiableProperties verifiableProperties)
{
}