tomcat - Apache Spark 导致 Tomcat 正常关闭

Question

Tomcat 版本:7.0.47

我有一个使用 Apache Spark 的网络应用程序。我的 Web 应用程序充当 Apache Spark 驱动程序。

当远程独立 spark 集群不可用时，Spark 上下文将关闭并显示日志 org.apache.spark.util.Utils - Shutdown hook called

越早发生这种情况，Tomcat 也开始正常关闭。我在tomcat中能看到的唯一日志是[exec] Result: 50

当 spark 调用 it's shut down hook 时，Tomcat 关闭的原因是什么？

Spark 日志

SLF4J: Class path contains multiple SLF4J bindings.
SLF4J: Found binding in [jar:file:/data/downloads/spark-1.4.1-bin-hadoop2.6/lib/spark-assembly-1.4.1-hadoop2.6.0.jar!/org/slf4j/impl/StaticLoggerBinder.class]
SLF4J: Found binding in [jar:file:/data/downloads/spark-1.4.1-bin-hadoop2.6/lib/spark-examples-1.4.1-hadoop2.6.0.jar!/org/slf4j/impl/StaticLoggerBinder.class]
SLF4J: See http://www.slf4j.org/codes.html#multiple_bindings for an explanation.
SLF4J: Actual binding is of type [org.slf4j.impl.Log4jLoggerFactory]
Using Spark's default log4j profile: org/apache/spark/log4j-defaults.properties
15/12/09 17:11:23 INFO SparkContext: Running Spark version 1.4.1
15/12/09 17:11:24 WARN NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
15/12/09 17:11:24 WARN Utils: Your hostname, pesamara-mobl-vm1 resolves to a loopback address: 127.0.0.1; using 10.30.9.107 instead (on interface eth0)
15/12/09 17:11:24 WARN Utils: Set SPARK_LOCAL_IP if you need to bind to another address
15/12/09 17:11:25 INFO SecurityManager: Changing view acls to: pes
15/12/09 17:11:25 INFO SecurityManager: Changing modify acls to: pes
15/12/09 17:11:25 INFO SecurityManager: SecurityManager: authentication disabled; ui acls disabled; users with view permissions: Set(pes); users with modify permissions: Set(pes)
15/12/09 17:11:26 INFO Slf4jLogger: Slf4jLogger started
15/12/09 17:11:27 INFO Remoting: Starting remoting
15/12/09 17:11:27 INFO Remoting: Remoting started; listening on addresses :[akka.tcp://sparkDriver@10.30.9.107:55740]
15/12/09 17:11:27 INFO Utils: Successfully started service 'sparkDriver' on port 55740.
15/12/09 17:11:27 INFO SparkEnv: Registering MapOutputTracker
15/12/09 17:11:27 INFO SparkEnv: Registering BlockManagerMaster
15/12/09 17:11:27 INFO DiskBlockManager: Created local directory at /tmp/spark-30d61b03-0b1c-4250-b68e-c2404c7884a8/blockmgr-3226ed7e-f8e5-40a2-bfb1-ffabb51cd0e0
15/12/09 17:11:28 INFO MemoryStore: MemoryStore started with capacity 491.5 MB
15/12/09 17:11:28 INFO HttpFileServer: HTTP File server directory is /tmp/spark-30d61b03-0b1c-4250-b68e-c2404c7884a8/httpd-7f2572c2-5677-446e-a80a-6f9d05ee2891
15/12/09 17:11:28 INFO HttpServer: Starting HTTP Server
15/12/09 17:11:28 INFO Utils: Successfully started service 'HTTP file server' on port 45047.
15/12/09 17:11:28 INFO SparkEnv: Registering OutputCommitCoordinator
15/12/09 17:11:28 INFO Utils: Successfully started service 'SparkUI' on port 4040.
15/12/09 17:11:28 INFO SparkUI: Started SparkUI at http://10.30.9.107:4040
15/12/09 17:11:29 INFO FairSchedulableBuilder: Created default pool default, schedulingMode: FIFO, minShare: 0, weight: 1
15/12/09 17:11:29 INFO AppClient$ClientActor: Connecting to master akka.tcp://sparkMaster@localhost2:7077/user/Master...
15/12/09 17:11:29 WARN AppClient$ClientActor: Could not connect to akka.tcp://sparkMaster@localhost2:7077: akka.remote.InvalidAssociation: Invalid address: akka.tcp://sparkMaster@localhost2:7077
15/12/09 17:11:29 WARN Remoting: Tried to associate with unreachable remote address [akka.tcp://sparkMaster@localhost2:7077]. Address is now gated for 5000 ms, all messages to this address will be delivered to dead letters. Reason: localhost2: unknown error
15/12/09 17:11:49 INFO AppClient$ClientActor: Connecting to master akka.tcp://sparkMaster@localhost2:7077/user/Master...
15/12/09 17:11:49 WARN AppClient$ClientActor: Could not connect to akka.tcp://sparkMaster@localhost2:7077: akka.remote.InvalidAssociation: Invalid address: akka.tcp://sparkMaster@localhost2:7077
15/12/09 17:11:49 WARN Remoting: Tried to associate with unreachable remote address [akka.tcp://sparkMaster@localhost2:7077]. Address is now gated for 5000 ms, all messages to this address will be delivered to dead letters. Reason: localhost2: unknown error
15/12/09 17:12:09 INFO AppClient$ClientActor: Connecting to master akka.tcp://sparkMaster@localhost2:7077/user/Master...
15/12/09 17:12:09 WARN AppClient$ClientActor: Could not connect to akka.tcp://sparkMaster@localhost2:7077: akka.remote.InvalidAssociation: Invalid address: akka.tcp://sparkMaster@localhost2:7077
15/12/09 17:12:09 WARN Remoting: Tried to associate with unreachable remote address [akka.tcp://sparkMaster@localhost2:7077]. Address is now gated for 5000 ms, all messages to this address will be delivered to dead letters. Reason: localhost2: unknown error
15/12/09 17:12:29 ERROR SparkDeploySchedulerBackend: Application has been killed. Reason: All masters are unresponsive! Giving up.
15/12/09 17:12:29 WARN SparkDeploySchedulerBackend: Application ID is not initialized yet.
15/12/09 17:12:29 INFO SparkUI: Stopped Spark web UI at http://10.30.9.107:4040
15/12/09 17:12:29 INFO DAGScheduler: Stopping DAGScheduler
15/12/09 17:12:29 INFO SparkDeploySchedulerBackend: Shutting down all executors
15/12/09 17:12:29 INFO SparkDeploySchedulerBackend: Asking each executor to shut down
15/12/09 17:12:29 ERROR OneForOneStrategy: 
java.lang.NullPointerException
    at org.apache.spark.deploy.client.AppClient$ClientActor$$anonfun$receiveWithLogging$1.applyOrElse(AppClient.scala:160)
    at scala.runtime.AbstractPartialFunction$mcVL$sp.apply$mcVL$sp(AbstractPartialFunction.scala:33)
    at scala.runtime.AbstractPartialFunction$mcVL$sp.apply(AbstractPartialFunction.scala:33)
    at scala.runtime.AbstractPartialFunction$mcVL$sp.apply(AbstractPartialFunction.scala:25)
    at org.apache.spark.util.ActorLogReceive$$anon$1.apply(ActorLogReceive.scala:59)
    at org.apache.spark.util.ActorLogReceive$$anon$1.apply(ActorLogReceive.scala:42)
    at scala.PartialFunction$class.applyOrElse(PartialFunction.scala:118)
    at org.apache.spark.util.ActorLogReceive$$anon$1.applyOrElse(ActorLogReceive.scala:42)
    at akka.actor.Actor$class.aroundReceive(Actor.scala:465)
    at org.apache.spark.deploy.client.AppClient$ClientActor.aroundReceive(AppClient.scala:61)
    at akka.actor.ActorCell.receiveMessage(ActorCell.scala:516)
    at akka.actor.ActorCell.invoke(ActorCell.scala:487)
    at akka.dispatch.Mailbox.processMailbox(Mailbox.scala:238)
    at akka.dispatch.Mailbox.run(Mailbox.scala:220)
    at akka.dispatch.ForkJoinExecutorConfigurator$AkkaForkJoinTask.exec(AbstractDispatcher.scala:393)
    at scala.concurrent.forkjoin.ForkJoinTask.doExec(ForkJoinTask.java:260)
    at scala.concurrent.forkjoin.ForkJoinPool$WorkQueue.runTask(ForkJoinPool.java:1339)
    at scala.concurrent.forkjoin.ForkJoinPool.runWorker(ForkJoinPool.java:1979)
    at scala.concurrent.forkjoin.ForkJoinWorkerThread.run(ForkJoinWorkerThread.java:107)
15/12/09 17:12:29 INFO AppClient$ClientActor: Connecting to master akka.tcp://sparkMaster@localhost2:7077/user/Master...
15/12/09 17:12:29 WARN AppClient$ClientActor: Could not connect to akka.tcp://sparkMaster@localhost2:7077: akka.remote.InvalidAssociation: Invalid address: akka.tcp://sparkMaster@localhost2:7077
15/12/09 17:12:29 WARN Remoting: Tried to associate with unreachable remote address [akka.tcp://sparkMaster@localhost2:7077]. Address is now gated for 5000 ms, all messages to this address will be delivered to dead letters. Reason: localhost2: unknown error
15/12/09 17:12:29 INFO Utils: Successfully started service 'org.apache.spark.network.netty.NettyBlockTransferService' on port 54184.
15/12/09 17:12:29 INFO NettyBlockTransferService: Server created on 54184
15/12/09 17:12:29 INFO BlockManagerMaster: Trying to register BlockManager
15/12/09 17:12:29 INFO BlockManagerMasterEndpoint: Registering block manager 10.30.9.107:54184 with 491.5 MB RAM, BlockManagerId(driver, 10.30.9.107, 54184)
15/12/09 17:12:29 INFO BlockManagerMaster: Registered BlockManager
15/12/09 17:12:30 ERROR SparkContext: Error initializing SparkContext.
java.lang.IllegalStateException: Cannot call methods on a stopped SparkContext
    at org.apache.spark.SparkContext.org$apache$spark$SparkContext$$assertNotStopped(SparkContext.scala:103)
    at org.apache.spark.SparkContext.getSchedulingMode(SparkContext.scala:1503)
    at org.apache.spark.SparkContext.postEnvironmentUpdate(SparkContext.scala:2007)
    at org.apache.spark.SparkContext.<init>(SparkContext.scala:543)
    at org.apache.spark.api.java.JavaSparkContext.<init>(JavaSparkContext.scala:61)
    at org.apache.spark.examples.sql.SparkContextTest.main(SparkContextTest.java:32)
15/12/09 17:12:30 INFO SparkContext: SparkContext already stopped.
java.lang.IllegalStateException: Cannot call methods on a stopped SparkContext
    at org.apache.spark.SparkContext.org$apache$spark$SparkContext$$assertNotStopped(SparkContext.scala:103)
    at org.apache.spark.SparkContext.getSchedulingMode(SparkContext.scala:1503)
    at org.apache.spark.SparkContext.postEnvironmentUpdate(SparkContext.scala:2007)
    at org.apache.spark.SparkContext.<init>(SparkContext.scala:543)
    at org.apache.spark.api.java.JavaSparkContext.<init>(JavaSparkContext.scala:61)
    at org.apache.spark.examples.sql.SparkContextTest.main(SparkContextTest.java:32)
15/12/09 17:12:30 INFO DiskBlockManager: Shutdown hook called
15/12/09 17:12:30 INFO Utils: path = /tmp/spark-30d61b03-0b1c-4250-b68e-c2404c7884a8/blockmgr-3226ed7e-f8e5-40a2-bfb1-ffabb51cd0e0, already present as root for deletion.
15/12/09 17:12:30 INFO Utils: Shutdown hook called
15/12/09 17:12:30 INFO Utils: Deleting directory /tmp/spark-30d61b03-0b1c-4250-b68e-c2404c7884a8/httpd-7f2572c2-5677-446e-a80a-6f9d05ee2891
15/12/09 17:12:30 INFO Utils: Deleting directory /tmp/spark-30d61b03-0b1c-4250-b68e-c2404c7884a8

Answer 1

如果您使用 IP 地址代替主机名，则在 SparkConf 中设置主机时使用 spark://hostname:7077 作为 spark 主机。

我遇到了同样的问题，通过在 spark master 中使用主机名而不是使用 IP 地址解决了这个问题。

问候，火锅