如何将JMX远程连接到Dataproc上的Spark worker [英] How to connect with JMX remotely to Spark worker on Dataproc

问题描述

通过添加以下内容，我可以很好地连接到驱动程序:

I can connect to the driver just fine by adding the following:

spark.driver.extraJavaOptions=-Dcom.sun.management.jmxremote \
                              -Dcom.sun.management.jmxremote.port=9178 \
                              -Dcom.sun.management.jmxremote.authenticate=false \
                              -Dcom.sun.management.jmxremote.ssl=false

但是做...

spark.executor.extraJavaOptions=-Dcom.sun.management.jmxremote \
                                -Dcom.sun.management.jmxremote.port=9178 \
                                -Dcom.sun.management.jmxremote.authenticate=false \
                                -Dcom.sun.management.jmxremote.ssl=false

...只会在驱动程序上产生一堆错误...

... only yield a bunch of errors on the driver ...

Container id: container_1501548048292_0024_01_000003
Exit code: 1
Stack trace: ExitCodeException exitCode=1: 
    at org.apache.hadoop.util.Shell.runCommand(Shell.java:972)
    at org.apache.hadoop.util.Shell.run(Shell.java:869)
    at org.apache.hadoop.util.Shell$ShellCommandExecutor.execute(Shell.java:1170)
    at org.apache.hadoop.yarn.server.nodemanager.DefaultContainerExecutor.launchContainer(DefaultContainerExecutor.java:236)
    at org.apache.hadoop.yarn.server.nodemanager.containermanager.launcher.ContainerLaunch.call(ContainerLaunch.java:305)
    at org.apache.hadoop.yarn.server.nodemanager.containermanager.launcher.ContainerLaunch.call(ContainerLaunch.java:84)
    at java.util.concurrent.FutureTask.run(FutureTask.java:266)
    at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
    at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
    at java.lang.Thread.run(Thread.java:748)


Container exited with a non-zero exit code 1

...最终使工作崩溃.

... and finally crashes the job.

工人没有任何错误，只需退出:

There are no errors on the workers, it simply exits with:

[org.apache.spark.util.ShutdownHookManager] - Shutdown hook called

Spark v2.2.0，并且该群集是一个简单的1m-2w配置，并且我的作业运行时没有执行者参数而没有问题.

Spark v2.2.0, and the cluster is a simple 1m-2w-configuration, and my jobs run without issues without the executor parameters.

推荐答案

正如 Rick Mortiz 指出的那样，问题是执行器jmx的端口冲突.

As Rick Mortiz pointed out, the issue was conflicting ports for the executor jmx.

设置:

-Dcom.sun.management.jmxremote.port=0

产生一个随机端口，并从Spark中删除了错误.要使用do找出最终使用哪个端口，请执行以下操作:

yields a random port, and removed the errors from Spark. To figure out which port it ends up using do:

netstat -alp | grep LISTEN.*<executor-pid>/java

列出了该进程当前打开的端口.

which lists the currently open ports for that process.

这篇关于如何将JMX远程连接到Dataproc上的Spark worker的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！