【整合】Flink提交运行中常见问题总结

一、提交jar到Flink集群时候出现异常

提交jar到Flink集群时候出现：
java.util.ServiceConfigurationError: org.apache.hadoop.fs.FileSystem: Provider org.apache.hadoop.fs.viewfs.ViewFileSystem could not be instantiated

可能原因解释：

出现该问题，往往是由于没有争取读取到hadoop中配置信息，

解决方法：

HADOOP_HOME=/hadoop/hadoop-2.7.2 
HADOOP_CONF_DIR=/hadoop/hadoop-2.7.2/etc/hadoop/ 
FLINK_HOME=/home/flink-1.4.2

同时建议使用Flink提供的官方flink-hadoop依赖包，自身提供的hadoop的jar包坑不全
在maven中pom.xml中加入如下依赖
使用不同的flink官方包：如果要做checkpoint功能，需要此时需要配置

 
        
            org.apache.flink
            flink-shaded-hadoop2
            1.4.2
        

ps:其他的maven中的依赖项，见我的另外一篇博客：
https://blog.csdn.net/fct2001140269/article/details/84864151

二、提交执行Flink的jar包时候，出现错误[Flink JobExecutionException: akka.client.timeout]

提交执行Flink的jar包时候，出现错误[Flink JobExecutionException: akka.client.timeout]

问题分析：

表面来看就是你的工程没有在规定的时间内（Flink集群默认配置中是60s），可能是你的工程项目比较大，jobManager进程难以在规定的短时间内完成Flink的逻辑topology图的构建，或者难以在规定的时间内，完成各个算子（例如：richMapFunction()）的初始化等工作；

解决办法：

建议在集群中配置如下选型：增加job作业提交允许的构建和初始化读取资源的时间。
 

    -akka.client.timeout:600s
    -akka.ask.timeout:600s

重新提交jar包到集群中，等待一段时间可以在web-ui上查看到任务提交情况。（如果失败，可能还有其他原因导致到异常！）

三、启动不起来

查看JobManager日志：

WARN  org.apache.flink.runtime.webmonitor.JobManagerRetriever       - Failed to retrieve leader gateway and port.
akka.actor.ActorNotFound: Actor not found for: ActorSelection[Anchor(akka.tcp://flink@t-sha1-flk-01:6123/), Path(/user/jobmanager)]
    at akka.actor.ActorSelection$$anonfun$resolveOne$1.apply(ActorSelection.scala:65)
    at akka.actor.ActorSelection$$anonfun$resolveOne$1.apply(ActorSelection.scala:63)
    at scala.concurrent.impl.CallbackRunnable.run(Promise.scala:32)
    at akka.dispatch.BatchingExecutor$AbstractBatch.processBatch(BatchingExecutor.scala:55)
    at akka.dispatch.BatchingExecutor$Batch.run(BatchingExecutor.scala:73)
    at akka.dispatch.ExecutionContexts$sameThreadExecutionContext$.unbatchedExecute(Future.scala:74)
    at akka.dispatch.BatchingExecutor$class.execute(BatchingExecutor.scala:120)
    at akka.dispatch.ExecutionContexts$sameThreadExecutionContext$.execute(Future.scala:73)
    at scala.concurrent.impl.CallbackRunnable.executeWithValue(Promise.scala:40)
    at scala.concurrent.impl.Promise$DefaultPromise.tryComplete(Promise.scala:248)
    at akka.pattern.PromiseActorRef$$anonfun$1.apply$mcV$sp(AskSupport.scala:334)
    at akka.actor.Scheduler$$anon$7.run(Scheduler.scala:117)
    at scala.concurrent.Future$InternalCallbackExecutor$.unbatchedExecute(Future.scala:599)
    at scala.concurrent.BatchingExecutor$class.execute(BatchingExecutor.scala:109)
    at scala.concurrent.Future$InternalCallbackExecutor$.execute(Future.scala:597)
    at akka.actor.LightArrayRevolverScheduler$TaskHolder.executeTask(Scheduler.scala:474)
    at akka.actor.LightArrayRevolverScheduler$$anon$8.executeBucket$1(Scheduler.scala:425)
    at akka.actor.LightArrayRevolverScheduler$$anon$8.nextTick(Scheduler.scala:429)
    at akka.actor.LightArrayRevolverScheduler$$anon$8.run(Scheduler.scala:381)
    at java.lang.Thread.run(Thread.java:748)

 解决方案：

/etc/hosts中配置的主机名都是小写，但是在Flink配置文件（flink-config.yaml、masters、slaves）中配置的都是大写的hostname，将flink配置文件中的hostname都改为小写或者IP地址。

四、运行一段时间退出

AsynchronousException{java.lang.Exception: Could not materialize checkpoint 4 for operator Compute By Event Time (1/12).}
    at org.apache.flink.streaming.runtime.tasks.StreamTask$AsyncCheckpointRunnable.run(StreamTask.java:970)
    at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
    at java.util.concurrent.FutureTask.run(FutureTask.java:266)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
    at java.lang.Thread.run(Thread.java:748)
Caused by: java.lang.Exception: Could not materialize checkpoint 4 for operator Compute By Event Time (1/12).
    ... 6 more
Caused by: java.util.concurrent.ExecutionException: java.io.IOException: Size of the state is larger than the maximum permitted memory-backed state. Size=7061809 , maxSize=5242880 . Consider using a different state backend, like the File System State backend.
    at java.util.concurrent.FutureTask.report(FutureTask.java:122)
    at java.util.concurrent.FutureTask.get(FutureTask.java:192)
    at org.apache.flink.util.FutureUtil.runIfNotDoneAndGet(FutureUtil.java:43)
    at org.apache.flink.streaming.runtime.tasks.StreamTask$AsyncCheckpointRunnable.run(StreamTask.java:897)
    ... 5 more
    Suppressed: java.lang.Exception: Could not properly cancel managed keyed state future.
        at org.apache.flink.streaming.api.operators.OperatorSnapshotResult.cancel(OperatorSnapshotResult.java:90)
        at org.apache.flink.streaming.runtime.tasks.StreamTask$AsyncCheckpointRunnable.cleanup(StreamTask.java:1023)
        at org.apache.flink.streaming.runtime.tasks.StreamTask$AsyncCheckpointRunnable.run(StreamTask.java:961)
        ... 5 more
    Caused by: java.util.concurrent.ExecutionException: java.io.IOException: Size of the state is larger than the maximum permitted memory-backed state. Size=7061809 , maxSize=5242880 . Consider using a different state backend, like the File System State backend.
        at java.util.concurrent.FutureTask.report(FutureTask.java:122)
        at java.util.concurrent.FutureTask.get(FutureTask.java:192)
        at org.apache.flink.util.FutureUtil.runIfNotDoneAndGet(FutureUtil.java:43)
        at org.apache.flink.runtime.state.StateUtil.discardStateFuture(StateUtil.java:85)
        at org.apache.flink.streaming.api.operators.OperatorSnapshotResult.cancel(OperatorSnapshotResult.java:88)
        ... 7 more
    Caused by: java.io.IOException: Size of the state is larger than the maximum permitted memory-backed state. Size=7061809 , maxSize=5242880 . Consider using a different state backend, like the File System State backend.
        at org.apache.flink.runtime.state.memory.MemCheckpointStreamFactory.checkSize(MemCheckpointStreamFactory.java:64)
        at org.apache.flink.runtime.state.memory.MemCheckpointStreamFactory$MemoryCheckpointOutputStream.closeAndGetBytes(MemCheckpointStreamFactory.java:144)
        at org.apache.flink.runtime.state.memory.MemCheckpointStreamFactory$MemoryCheckpointOutputStream.closeAndGetHandle(MemCheckpointStreamFactory.java:125)
        at org.apache.flink.runtime.checkpoint.AbstractAsyncSnapshotIOCallable.closeStreamAndGetStateHandle(AbstractAsyncSnapshotIOCallable.java:100)
        at org.apache.flink.runtime.state.heap.HeapKeyedStateBackend$1.performOperation(HeapKeyedStateBackend.java:351)
        at org.apache.flink.runtime.state.heap.HeapKeyedStateBackend$1.performOperation(HeapKeyedStateBackend.java:329)
        at org.apache.flink.runtime.io.async.AbstractAsyncIOCallable.call(AbstractAsyncIOCallable.java:72)
        at java.util.concurrent.FutureTask.run(FutureTask.java:266)
        at org.apache.flink.runtime.state.heap.HeapKeyedStateBackend.snapshot(HeapKeyedStateBackend.java:372)
        at org.apache.flink.streaming.api.operators.AbstractStreamOperator.snapshotState(AbstractStreamOperator.java:397)
        at org.apache.flink.streaming.runtime.tasks.StreamTask$CheckpointingOperation.checkpointStreamOperator(StreamTask.java:1162)
        at org.apache.flink.streaming.runtime.tasks.StreamTask$CheckpointingOperation.executeCheckpointing(StreamTask.java:1094)
        at org.apache.flink.streaming.runtime.tasks.StreamTask.checkpointState(StreamTask.java:654)
        at org.apache.flink.streaming.runtime.tasks.StreamTask.performCheckpoint(StreamTask.java:590)
        at org.apache.flink.streaming.runtime.tasks.StreamTask.triggerCheckpointOnBarrier(StreamTask.java:543)
        at org.apache.flink.streaming.runtime.io.BarrierBuffer.notifyCheckpoint(BarrierBuffer.java:378)
        at org.apache.flink.streaming.runtime.io.BarrierBuffer.processBarrier(BarrierBuffer.java:281)
        at org.apache.flink.streaming.runtime.io.BarrierBuffer.getNextNonBlocked(BarrierBuffer.java:183)
        at org.apache.flink.streaming.runtime.io.StreamInputProcessor.processInput(StreamInputProcessor.java:213)
        at org.apache.flink.streaming.runtime.tasks.OneInputStreamTask.run(OneInputStreamTask.java:69)
        at org.apache.flink.streaming.runtime.tasks.StreamTask.invoke(StreamTask.java:263)
        at org.apache.flink.runtime.taskmanager.Task.run(Task.java:702)
        ... 1 more
    [CIRCULAR REFERENCE:java.io.IOException: Size of the state is larger than the maximum permitted memory-backed state. Size=7061809 , maxSize=5242880 . Consider using a different state backend, like the File System State backend.]

 解决方案：

   状态存储，默认是在内存中，改为存储到HDFS中：

state.backend.fs.checkpointdir: hdfs://t-sha1-flk-01:9000/flink-checkpoints

五、长时间运行后，多次重启

AsynchronousException{java.lang.Exception: Could not materialize checkpoint 1488 for operator Compute By Event Time -> (MonitorData, MonitorDataMapping, MonitorSamplingData) (6/6).}
	at org.apache.flink.streaming.runtime.tasks.StreamTaskAsyncCheckpointRunnable.run(StreamTask.java:948)atjava.util.concurrent.Executors'>[Math Processing Error]AsyncCheckpointRunnable.run(StreamTask.java:948)atjava.util.concurrent.ExecutorsAsyncCheckpointRunnable.run(StreamTask.java:948) 	at java.util.concurrent.ExecutorsRunnableAdapter.call(Executors.java:511)
	at java.util.concurrent.FutureTask.run(FutureTask.java:266)
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
	at java.lang.Thread.run(Thread.java:748)
Caused by: java.lang.Exception: Could not materialize checkpoint 1488 for operator Compute By Event Time -> (MonitorData, MonitorDataMapping, MonitorSamplingData) (6/6).
	... 6 more
Caused by: java.util.concurrent.ExecutionException: org.apache.hadoop.ipc.RemoteException(java.io.IOException): File /flink-checkpoints/8c274785f1ab027e6146a59364be645f/chk-1488/2c612f30-c57d-4ede-9025-9554ca11fd12 could only be replicated to 0 nodes instead of minReplication (=1).  There are 3 datanode(s) running and no node(s) are excluded in this operation.
	at org.apache.hadoop.hdfs.server.blockmanagement.BlockManager.chooseTarget4NewBlock(BlockManager.java:1628)
	at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getNewBlockTargets(FSNamesystem.java:3121)
	at org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getAdditionalBlock(FSNamesystem.java:3045)
	at org.apache.hadoop.hdfs.server.namenode.NameNodeRpcServer.addBlock(NameNodeRpcServer.java:725)
	at org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolServerSideTranslatorPB.addBlock(ClientNamenodeProtocolServerSideTranslatorPB.java:493)
	at org.apache.hadoop.hdfs.protocol.proto.ClientNamenodeProtocolProtosClientNamenodeProtocol'>[Math Processing Error]ClientNamenodeProtocolClientNamenodeProtocol2.callBlockingMethod(ClientNamenodeProtocolProtos.java)
	at org.apache.hadoop.ipc.ProtobufRpcEngineServer'>[Math Processing Error]ServerServerProtoBufRpcInvoker.call(ProtobufRpcEngine.java:616)
	at org.apache.hadoop.ipc.RPCServer.call(RPC.java:982)atorg.apache.hadoop.ipc.Server'>[Math Processing Error]Server.call(RPC.java:982)atorg.apache.hadoop.ipc.ServerServer.call(RPC.java:982) 	at org.apache.hadoop.ipc.ServerHandler1.run(Server.java:2217)atorg.apache.hadoop.ipc.Server'>[Math Processing Error]1.run(Server.java:2217)atorg.apache.hadoop.ipc.Server1.run(Server.java:2217) 	at org.apache.hadoop.ipc.ServerHandler1.run(Server.java:2213)atjava.security.AccessController.doPrivileged(NativeMethod)atjavax.security.auth.Subject.doAs(Subject.java:422)atorg.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1746)atorg.apache.hadoop.ipc.Server'>[Math Processing Error]1.run(Server.java:2213)atjava.security.AccessController.doPrivileged(NativeMethod)atjavax.security.auth.Subject.doAs(Subject.java:422)atorg.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1746)atorg.apache.hadoop.ipc.Server1.run(Server.java:2213) 	at java.security.AccessController.doPrivileged(Native Method) 	at javax.security.auth.Subject.doAs(Subject.java:422) 	at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1746) 	at org.apache.hadoop.ipc.ServerHandler.run(Server.java:2213)

 查看hdfs日志：

WARN org.apache.hadoop.hdfs.protocol.BlockStoragePolicy: 
Failed to place enough replicas: expected size is 2 but only 0 storage types can be selected 
(replication=3, selected=[], unavailable=[DISK], removed=[DISK, DISK],
 policy=BlockStoragePolicy{HOT:7, storageTypes=[DISK], creationFallbacks=[], replicationFallbacks=[ARCHIVE]})

  搭建的Flink使用HDFS作为CheckPoint的存储，当flink重启时，原来的checkpoint没有用了，我就手动给删了，不知道和这个有没有关系，为了不继续报异常，便重启了Flink、HDFS，重启后不再有异常信息了。

   但是查看HDFS日志时，发现如下警告（不合规范的URI格式）：

WARN org.apache.hadoop.hdfs.server.common.Util:
Path /mnt/hadoop/dfs/name should be specified as a URI in configuration files. 
Please update hdfs configuration

   原来是配置错了，hdfs-site.xml中的

    
      dfs.namenode.name.dir
      /mnt/hadoop/dfs/name
    

应该改为：

    
      dfs.namenode.name.dir
      file:/mnt/hadoop/dfs/name
    

  至此问题解决，根上的问题应该是hdfs-site.xml配置的不对导致的。 

六、Unable to load native-hadoop library for your platform

Flink启动时，有时会有如下警告信息：

WARN  org.apache.hadoop.util.NativeCodeLoader    

- Unable to load native-hadoop library for your platform... 

using builtin-java classes where applicable

  参考资料1：http://blog.csdn.net/jack85986370/article/details/51902871

  解决方案：

编辑/etc/profile文件，增加

export HADOOP_COMMON_LIB_NATIVE_DIR=$HADOOP_HOME/lib/native  

export HADOOP_OPTS="-Djava.library.path=$HADOOP_HOME/lib"  

  未能解决该问题

七、hadoop checknative -a 

WARN bzip2.Bzip2Factory: Failed to load/initialize native-bzip2 library system-native, will use pure-Java version
INFO zlib.ZlibFactory: Successfully loaded & initialized native-zlib library
Native library checking:
hadoop:  true /usr/hadoop-2.7.3/lib/native/libhadoop.so.1.0.0
zlib:    true /lib64/libz.so.1
snappy:  false 
lz4:     true revision:99
bzip2:   false 
openssl: false Cannot load libcrypto.so (libcrypto.so: cannot open shared object file: No such file or directory)!
INFO util.ExitUtil: Exiting with status 1

    参考资料：http://blog.csdn.net/zhangzhaokun/article/details/50951238

解决方案：

 cd /usr/lib64/
 ln -s libcrypto.so.1.0.1e libcrypto.so

八、TaskManager退出

   Flink运行一段时间后，出现TaskManager退出情况，通过jvisualvm抓取TaskManager的Dump，使用MAT进行分析，结果如下：

One instance of "org.apache.flink.runtime.io.network.buffer.NetworkBufferPool"
loaded by "sun.misc.Launcher$AppClassLoader @ 0x6c01de310" occupies 403,429,704 (76.24%) bytes. 
The memory is accumulated in one instance of "java.lang.Object[]" loaded by "".

Keywords
sun.misc.Launcher$AppClassLoader @ 0x6c01de310
java.lang.Object[]
org.apache.flink.runtime.io.network.buffer.NetworkBufferPool

 发现是网络缓冲池不足，查到一篇文章：

 https://issues.apache.org/jira/browse/FLINK-4536

 和我遇到的情况差不多，也是使用了InfluxDB作为Sink，最后在Close里进行关闭，问题解决。

 另外，在$FLINK_HOME/conf/flink-conf.yaml中，也有关于TaskManager网络栈的配置，暂时未调整。

# The number of buffers for the network stack.
#
# taskmanager.network.numberOfBuffers: 2048

九、Kafka partition leader切换导致Flink重启

现象：

 7.1 Flink重启，查看日志，显示：

java.lang.Exception: Failed to send data to Kafka: This server is not the leader for that topic-partition.
  at org.apache.flink.streaming.connectors.kafka.FlinkKafkaProducerBase.checkErroneous(FlinkKafkaProducerBase.java:373)
  at org.apache.flink.streaming.connectors.kafka.FlinkKafkaProducerBase.invoke(FlinkKafkaProducerBase.java:280)
  at org.apache.flink.streaming.api.operators.StreamSink.processElement(StreamSink.java:41)
  at org.apache.flink.streaming.runtime.io.StreamInputProcessor.processInput(StreamInputProcessor.java:206)
  at org.apache.flink.streaming.runtime.tasks.OneInputStreamTask.run(OneInputStreamTask.java:69)
  at org.apache.flink.streaming.runtime.tasks.StreamTask.invoke(StreamTask.java:263)
  at org.apache.flink.runtime.taskmanager.Task.run(Task.java:702)
  at java.lang.Thread.run(Thread.java:748)
Caused by: org.apache.kafka.common.errors.NotLeaderForPartitionException: This server is not the leader for that topic-partition.

7.2 查看Kafka的Controller日志，显示：

 INFO [SessionExpirationListener on 10], ZK expired; shut down all controller components and 

try to re-elect (kafka.controller.KafkaController$SessionExpirationListener)

7.3 设置retries参数

   参考：http://colabug.com/122248.html 以及 Kafka官方文档（http://kafka.apache.org/082/documentation.html#producerconfigs），关于producer参数设置

   设置了retries参数，可以在Kafka的Partition发生leader切换时，Flink不重启，而是做3次尝试：

        kafkaProducerConfig
          {
                "bootstrap.servers": "192.169.2.20:9093,192.169.2.21:9093,192.169.2.22:9093"
                "retries":3
          }