Running H2O Cluster in Background and at Specific Port Number

In a few of my previous blog, I discussed H2O vs Sparkling Water, Sparking Water Shell: Cloud size under 12 Exception, and Access Sparkling Water via R Studio.

When using H2O Sparkling Water, there are two common issues. The first one is that the default port number is 54321. However, if the H2O cluster has been bounced multiple times, the assigned port could be assigned to a different port number, actually next available port number, 54323. If the cluster is used by many data analysts, it is inconvenient to inform all of them every time the port number is changed. You want users remember only one port number.

The second issue is that the sparkling shell session can not be running in the background. If close the putty session running the sparkling shell, the H2O cluster is terminated.

This blog discusses the solution to work around the above two issues.

For port number, there are actually two parameters related: spark.ext.h2o.client.port.base and spark.ext.h2o.node.port.base. The spark.ext.h2o.client.port.base is the port number for H2O UI while spark.ext.h2o.node.port.base is the port used by H2O cluster internally for the communication among H2O nodes. Make sure these two port numbers should be different. Having these two are the same will cause issue. I also add for the name of my H2O cluster.

I created two separate scripts: for the running command for sparkling shell and sparkling-shell-init.scala for starting up commands for H2O cluster in scala.

[wzhou@enkbda1node05 ~]$ cat
bin/sparkling-shell \
--master yarn \
--conf \
--conf spark.ext.h2o.client.port.base=26000 \
--conf spark.ext.h2o.node.port.base=26005 \
--conf spark.executor.instances=10 \
--conf spark.executor.memory=12g \
--conf spark.driver.memory=8g \
--conf spark.executor.cores=4 \
--conf spark.yarn.executor.memoryOverhead=4g \
--conf spark.yarn.driver.memoryOverhead=4g \
--conf spark.scheduler.maxRegisteredResourcesWaitingTime=1000000 \
--conf \
--conf spark.dynamicAllocation.enabled=false \
--conf spark.sql.autoBroadcastJoinThreshold=-1 \
--conf spark.locality.wait=30000 \
--conf spark.yarn.queue=WZTestPool \
--conf spark.scheduler.minRegisteredResourcesRatio=1 \
-i  sparkling-shell-init.scala

The code for sparkling-shell-init.scala.

import org.apache.spark.h2o._
val h2oContext = H2OContext.getOrCreate(spark)
import h2oContext._

To execute sparkling-shell in background, my first try was to use nohup. It didn’t work. When calling sparkling-shell-init.scala script, it automatically adds :quit command at the end and terminate H2O cluster.

When I did work on Exadata, I used to use screen command a lot. It is a very useful tool for protecting long running critical job execution, like patching/upgrade and import/export. Therefore, I use the same trick in screen to help me to get around the background issue. Here are the steps.

1. Start screen session
Use screen -ls command to check whether I have screen available.

[wzhou@enkbda1node05 ~]$ screen -ls
No Sockets found in /var/run/screen/S-wzhou.

Start a screen session.
[wzhou@enkbda1node05 ~]$ screen

2. Start H2O Cluster
Run the script to start H2O cluster.

[wzhou@enkbda1node05 ~]$ ./

  Spark master (MASTER)     : yarn
  Spark home   (SPARK_HOME) :
  H2O build version         : (weierstrass)
  Spark build version       : 2.2.0
  Scala version             : 2.11

SLF4J: Class path contains multiple SLF4J bindings.
SLF4J: Found binding in [jar:file:/opt/anaconda2/lib/python2.7/site-packages/pyspark/jars/slf4j-log4j12-1.7.16.jar!/org/slf4j/impl/StaticLoggerBinder.class]
SLF4J: Found binding in [jar:file:/opt/cloudera/parcels/CDH-5.10.1-1.cdh5.10.1.p0.10/jars/slf4j-log4j12-1.7.5.jar!/org/slf4j/impl/StaticLoggerBinder.class]
SLF4J: See for an explanation.
SLF4J: Actual binding is of type [org.slf4j.impl.Log4jLoggerFactory]
Setting default log level to "WARN".
To adjust logging level use sc.setLogLevel(newLevel). For SparkR, use setLogLevel(newLevel).
17/12/09 10:12:58 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
17/12/09 10:12:58 WARN shortcircuit.DomainSocketFactory: The short-circuit local reads feature cannot be used because libhadoop cannot be loaded.
17/12/09 10:12:59 WARN yarn.Client: Neither spark.yarn.jars nor spark.yarn.archive is set, falling back to uploading libraries under SPARK_HOME.
Spark context Web UI available at
Spark context available as 'sc' (master = yarn, app id = application_2567590118914_1007).
Spark session available as 'spark'.
Loading sparkling-shell-init.scala...
import org.apache.spark.h2o._
h2oContext: org.apache.spark.h2o.H2OContext =

Sparkling Water Context:
 * H2O name: WeidongH2O-Cluster
 * cluster size: 10
 * list of used nodes:
  (executorId, host, port)

  Open H2O Flow in browser: (CMD + click in Mac OSX)

import h2oContext._

Welcome to
      ____              __
     / __/__  ___ _____/ /__
    _\ \/ _ \/ _ `/ __/  '_/
   /___/ .__/\_,_/_/ /_/\_\   version 2.2.0

Using Scala version 2.11.8 (Java HotSpot(TM) 64-Bit Server VM, Java 1.8.0_121)
Type in expressions to have them evaluated.
Type :help for more information.

3. Verify H2O Cluster
Right now, close the session. Open a new session. Check out the existing screen session.

[wzhou@enkbda1node05 ~]$ screen -ls
There is a screen on:
        19044.pts-0.enkbda1node05       (Detached)
1 Socket in /var/run/screen/S-wzhou.

Now, attach to the existing session
[wzhou@enkbda1node05 ~]$ screen -x 19044

You should see the original session that was running sparking shell. The UI is still working as expected.

Resolve Sparklyr not Respond Issue on Port 8880

Recently I was approached by one of my clients to help them to investigate a weird Sparklyr issue. sparklyr is an interface between R and Spark introduced by RStudio about a years ago. The following is the the sparklyr architecture.

When trying to do sc <- spark_connect in RStudio, we got two errors as follows:

  • Failed while connecting to sparklyr to port (8880) for sessionid (3859): Gateway in port (8880) did not respond.
  • Exception in thread “main” java.lang.NoClassDefFoundError: org/apache/hadoop/fs/FSDataInputStream
  • Here is the detail message.

    > library(sparklyr)
    > library(dplyr) 
    > sc <- spark_connect(master = "yarn-client", config=spark_config(), version="1.6.0", spark_home = '/opt/cloudera/parcels/CDH/lib/spark/')
    Error in force(code) :
    Failed while connecting to sparklyr to port (8880) for sessionid (3859): Gateway in port (8880) did not respond.
    Path: /opt/cloudera/parcels/CDH-5.10.1-1.cdh5.10.1.p0.10/lib/spark/bin/spark-submit
    Parameters: --class, sparklyr.Shell, --jars, '/usr/lib64/R/library/sparklyr/java/spark-csv_2.11-1.3.0.jar','/usr/lib64/R/library/sparklyr/java/commons-csv-1.1.jar','/usr/lib64/R/library/sparklyr/java/univocity-parsers-1.5.1.jar', '/usr/lib64/R/library/sparklyr/java/sparklyr-1.6-2.10.jar', 8880, 3859
    Log: /tmp/RtmzpSIMln/file9e23246605df7_spark.log
    Exception in thread "main" java.lang.NoClassDefFoundError: org/apache/hadoop/fs/FSDataInputStream
                   at org.apache.spark.deploy.SparkSubmitArguments.handle(SparkSubmitArguments.scala:394)
                   at org.apache.spark.launcher.SparkSubmitOptionParser.parse(
                   at org.apache.spark.deploy.SparkSubmitArguments.<init>(SparkSubmitArguments.scala:97)
                   at org.apache.spark.deploy.SparkSubmit$.main(SparkSubmit.scala:114)
                   at org.apache.spark.deploy.SparkSubmit.main(SparkSubmit.scala)
    Caused by: java.lang.ClassNotFoundException: org.apache.hadoop.fs.FSDataInputStream
                   at java.lang.ClassLoader.loadClass(
                   at sun.misc.Launcher$AppClassLoader.loadClass(
                   at java.lang.ClassLoader.loadClass(
                   ... 5 more

    Did some research and found many people having the similar issue. Ok, try their recommendations one by one as follows.

  • Set SPARK_HOME environment
  • Try run Sys.setEnv(SPARK_HOME = “/opt/cloudera/parcels/CDH/lib/spark/”). No, not working.

  • Install latest version sparklyr
  • My client installed sparklyr less than one month ago. I don’t see why this option makes sense. Don’t even pursue this path.

  • Check Java Installation
  • The R on the same server uses the same version of Java without any issue. I don’t see why Java installation become a major concern here. Ignore this one.

  • No Hadoop Installation
  • Someone said just Spark installation is not enough, not to have Hadoop Installation as well. Clearly it does not fit our situation. The server is an edge node and has hadoop installation.

  • Do not have a valid kerberos ticket
  • Running system2(‘klist’) does show no kerberos ticket. Ok, I then open up a shell within RStudio Server by clicking tools -> shell, then issuing the kinit command.
    Rerun system2(‘klist’) shows I have a valid kerberos ticket. Try again. still not working.
    Note: even it is not working, this step is necessary for further action when the issue is fixed. So still need to run this one no matter what the result is.

  • Create a different configure and pass to spark_connect
  • Someone recommended to create a new configure and pass it in. It looks like a good idea. Unfortunately, just doesn’t work.

    wzconfig <- spark_config()
    wzconfig$`` <- "client"
    wzconfig$spark.driver.cores <- 1
    wzconfig$spark.executor.cores <- 2
    wzconfig$spark.executor.memory <- "4G"
    sc <- spark_connect(master = "yarn-client", config=wzconfig, version="1.6.0", spark_home = '/opt/cloudera/parcels/CDH/lib/spark/')

    Actually this recommendation is missing another key parameter. By default the total number of executors launched is 2. I would usually bump up this number a little to get a better performance. You can use the following way to set up the
    total number of executors.

    wzconfig$spark.executor.instances <- 3

    Although this approach looks promising, still not working. But this approach is definitely a way to use for other purpose to better control the Spark resource usage.

  • Add remote address
  • Someone mentioned to set remote address. I thought this could another potential option as I resolved issues in Spark related to local IP issue in the past. So I add the following code in the configuration from the previous example, note parameter sparklyr.gateway.address is the hostname of active Resource Manager.

    wzconfig$sparklyr.gateway.remote <- TRUE
    wzconfig$sparklyr.gateway.address <- "" 

    Not working for this case.

  • Change deployment mode to yarn-cluster
  • This is probably the most unrealistic one. If connect as with master = “yarn-cluster”, the spark driver will be somewhere inside the Spark cluster. For our current case, I don’t believe this is the right solution. Don’t even try it.

  • Run Spark example
  • Someone recommended to run a spark-submit to verify SparkPi can be run from the environment. This looks reasonable. The good thing I figured out the issue before executing this one. But this definitely a valid and good test to verify spark-submit.

    /opt/cloudera/parcels/SPARK2/lib/spark2/bin/spark-submit --class org.apache.spark.examples.SparkPi --deploy-mode client --master yarn /opt/cloudera/parcels/SPARK2/lib/spark2/examples/jars/spark-examples_2.11-2.1.0.jar 10
  • HA for yarn-cluster
  • There is an interesting post Add support for `yarn-cluster` with high availability #905 discussing about the issue might relate to multiple resource managers. We use HA and this post is an interesting one. But might not fit into our case because I feel we have not reached to the HA part yet with Class Not Found message.

  • Need to set JAVA_HOME
  • Verified it and we have it. So this is not the issue.

  • My Solution
  • After reviewing or trying out some of above solutions, I like to go back my way of thinking. I must say I am not an expert in R or RStudio with very limited knowledge about how it works. But I did have extensive background in Spark tuning and trouble shooting.

    I know the error message Gateway in port (8880) did not respond is always the first message shows up and looks like the cause of the issue. But I thought differently. I believe the 2nd error NoClassDefFoundError: org/apache/hadoop/fs/FSDataInputStream looks more suspicious than the first one. Early this year I helped one of another clients on a weird Spark job issue, which is in the end, was caused by the incorrect path. It seems to me the path might not be right and cause Spark issue, then caused the first error of port not respond.

    With this idea in mind, I focused more the path verification. Run the command Sys.getenv() to get the environment as follows.

    > Sys.getenv()
    DISPLAY                 :0
    EDITOR                  vi
    GIT_ASKPASS             rpostback-askpass
    HADOOP_CONF_DIR         /etc/hadoop/conf.cloudera.hdfs
    HADOOP_HOME             /opt/cloudera/parcels/CDH
    HOME                    /home/wzhou
    JAVA_HOME               /usr/java/jdk1.8.0_144/jre
    LANG                    en_US.UTF-8
    LD_LIBRARY_PATH         /usr/lib64/R/lib::/lib:/usr/java/jdk1.8.0_92/jre/lib/amd64/server
    LN_S                    ln -s
    LOGNAME                 wzhou
    MAKE                    make
    PAGER                   /usr/bin/less
    PATH                    /usr/local/sbin:/usr/local/bin:/usr/bin:/usr/sbin:/sbin:/bin
    R_BROWSER               /usr/bin/xdg-open
    R_BZIPCMD               /usr/bin/bzip2
    R_DOC_DIR               /usr/share/doc/R-3.3.2
    R_GZIPCMD               /usr/bin/gzip
    R_HOME                  /usr/lib64/R
    R_INCLUDE_DIR           /usr/include/R
    R_LIBS_SITE             /usr/local/lib/R/site-library:/usr/local/lib/R/library:/usr/lib64/R/library:/usr/share/R/library
    R_LIBS_USER             ~/R/x86_64-redhat-linux-gnu-library/3.3
    R_PAPERSIZE             a4
    R_PDFVIEWER             /usr/bin/xdg-open
    R_PLATFORM              x86_64-redhat-linux-gnu
    R_PRINTCMD              lpr
    R_RD4PDF                times,hyper
    R_SESSION_TMPDIR        /tmp/RtmpZf9YMN
    R_SHARE_DIR             /usr/share/R
    R_SYSTEM_ABI            linux,gcc,gxx,gfortran,?
    R_TEXI2DVICMD           /usr/bin/texi2dvi
    R_UNZIPCMD              /usr/bin/unzip
    RS_RPOSTBACK_PATH       /usr/lib/rstudio-server/bin/rpostback
    RSTUDIO                 1
    RSTUDIO_PANDOC          /usr/lib/rstudio-server/bin/pandoc
    RSTUDIO_WINUTILS        bin/winutils
    SED                     /bin/sed
    SPARK_HOME              /opt/cloudera/parcels/SPARK2/lib/spark2
    SSH_ASKPASS             rpostback-askpass
    TAR                     /bin/gtar
    USER                    wzhou
    YARN_CONF_DIR           /etc/hadoop/conf.cloudera.yarn

    Ahhh, I noticed the environment missed SPARK_DIST_CLASSPATH environment variable. Then I set it using the command below just before sc <- spark_connect.

    Sys.setenv(SPARK_DIST_CLASSPATH = '/etc/hadoop/con:/opt/cloudera/parcels/CDH/lib/hadoop/libexec/../../hadoop/lib/*:/opt/cloudera/parcels/CDH/lib/hadoop/libexec/../../hadoop/.//*:/opt/cloudera/parcels/CDH/lib/hadoop/libexec/../../hadoop-hdfs/./:/opt/cloudera/parcels/CDH/lib/hadoop/libexec/../../hadoop-hdfs/lib/*:/opt/cloudera/parcels/CDH/lib/hadoop/libexec/../../hadoop-hdfs/.//*:/opt/cloudera/parcels/CDH/lib/hadoop/libexec/../../hadoop-yarn/lib/*:/opt/cloudera/parcels/CDH/lib/hadoop/libexec/../../hadoop-yarn/.//*:/opt/cloudera/parcels/CDH/lib/hadoop-mapreduce/lib/*:/opt/cloudera/parcels/CDH/lib/hadoop-mapreduce/.//*')

    Ok, try it again. Fantastic, it works!

    Ok, here is the real cause of the issue. It’s unnecessary to specify java path for sparklyr as it does not require a java path. However, it does have dependency on spark-submit. When spark-submit is executed, it can read java path and then submit the jar files to Spark accordingly. The cause of the issue if SPARK_DIST_CLASSPATH is not set, spark-submit is not working and Spark executors can not be launched.

    Other Note
    The following are some of useful commands:

    spark_home_dir() or spark_home
    config <- spark_config()


    Also there are a few useful articles about sparklyr and Rstudio:
    RStudio’s R Interface to Spark on Amazon EMR
    How to Install RStudio Server on CentOS 7
    Using R with Apache Spark
    sparklyr: a test drive on YARN
    Analyzing a billion NYC taxi trips in Spark