Spark SQL Thrift Server on CDH 5.3.0 - apache-spark

Question

I am trying to use CDH 5.3.0 to run Spark's Thrift Server. I'm trying to follow the Spark SQL instructions, but I can't even get the --help option to run successfully. In the output below, it dies because it can't find the HiveServer2 class.
$ /usr/lib/spark/sbin/start-thriftserver.sh --help
Usage./sbin/start-thriftserver [options] [thrift server options]
Options:
--master MASTER_URL spark://host:port, mesos://host:port, yarn, or local.
--deploy-mode DEPLOY_MODE Whether to launch the driver program locally ("client") or
on one of the worker machines inside the cluster ("cluster")
(Default: client).
--class CLASS_NAME Your application's main class (for Java / Scala apps).
--name NAME A name of your application.
--jars JARS Comma-separated list of local jars to include on the driver
and executor classpaths.
--py-files PY_FILES Comma-separated list of .zip, .egg, or .py files to place
on the PYTHONPATH for Python apps.
--files FILES Comma-separated list of files to be placed in the working
directory of each executor.
--conf PROP=VALUE Arbitrary Spark configuration property.
--properties-file FILE Path to a file from which to load extra properties. If not
specified, this will look for conf/spark-defaults.conf.
--driver-memory MEM Memory for driver (e.g. 1000M, 2G) (Default: 512M).
--driver-java-options Extra Java options to pass to the driver.
--driver-library-path Extra library path entries to pass to the driver.
--driver-class-path Extra class path entries to pass to the driver. Note that
jars added with --jars are automatically included in the
classpath.
--executor-memory MEM Memory per executor (e.g. 1000M, 2G) (Default: 1G).
--help, -h Show this help message and exit
--verbose, -v Print additional debug output
Spark standalone with cluster deploy mode only:
--driver-cores NUM Cores for driver (Default: 1).
--supervise If given, restarts the driver on failure.
Spark standalone and Mesos only:
--total-executor-cores NUM Total cores for all executors.
YARN-only:
--executor-cores NUM Number of cores per executor (Default: 1).
--queue QUEUE_NAME The YARN queue to submit to (Default: "default").
--num-executors NUM Number of executors to launch (Default: 2).
--archives ARCHIVES Comma separated list of archives to be extracted into the
working directory of each executor.
Thrift server options:
Exception in thread "main" java.lang.NoClassDefFoundError: org/apache/hive/service/server/HiveServer2
at java.lang.ClassLoader.defineClass1(Native Method)
at java.lang.ClassLoader.defineClass(ClassLoader.java:800)
at java.security.SecureClassLoader.defineClass(SecureClassLoader.java:142)
at java.net.URLClassLoader.defineClass(URLClassLoader.java:449)
at java.net.URLClassLoader.access$100(URLClassLoader.java:71)
at java.net.URLClassLoader$1.run(URLClassLoader.java:361)
at java.net.URLClassLoader$1.run(URLClassLoader.java:355)
at java.security.AccessController.doPrivileged(Native Method)
at java.net.URLClassLoader.findClass(URLClassLoader.java:354)
at java.lang.ClassLoader.loadClass(ClassLoader.java:425)
at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:308)
at java.lang.ClassLoader.loadClass(ClassLoader.java:358)
at sun.launcher.LauncherHelper.checkAndLoadMain(LauncherHelper.java:482)
Caused by: java.lang.ClassNotFoundException: org.apache.hive.service.server.HiveServer2
at java.net.URLClassLoader$1.run(URLClassLoader.java:366)
at java.net.URLClassLoader$1.run(URLClassLoader.java:355)
at java.security.AccessController.doPrivileged(Native Method)
at java.net.URLClassLoader.findClass(URLClassLoader.java:354)
at java.lang.ClassLoader.loadClass(ClassLoader.java:425)
at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:308)
at java.lang.ClassLoader.loadClass(ClassLoader.java:358)
... 13 more

Answer 1

As indicated by the error, the class is not in the classpath. Unfortunately, setting the CLASSPATH environment variable won't work. The only solution that I could find was to edit /usr/lib/spark/bin/compute-classpath.sh and add this line (it can go just about anywhere, but put it one line from the end to make it clear that it's an addition):
CLASSPATH="$CLASSPATH:/usr/lib/hive/lib/*"
Cloudera's release notes for 5.3.0 explicitly state "Spark SQL remains an experimental and unsupported feature in CDH", so it's not surprising that tweaks like this may be needed. Also, this response to a similar problem in CDH 5.2 suggests that the Hive jars are deliberately excluded by Cloudera for size reasons.

Answer 2

I have faced the same problem but I solved it in another way.
The cloudera CDH version was not 5.3.0 it was some version prior to that version so you will find the paths little different.
Simply the solution was to replace the spark-assembly-**.jar file that shipped with cloudera CDH by another version.
I downloaded spark from its official download page. The version I have downloaded was built for hadoop 2.4 and later. Extracting the downloaded file and look for spark-assembly-**.jar.
In the cloudera installation, I looked for the same file and I found it under that path /usr/lib/spark/libe/spark-assembly--.jar
The previous path actually was a symlink to the actual file. I uploaded the jar from spark download to the same path and make the symlink point to the new jar (ln -f -s target link).
Every thing works fine with me.

Answer 3

/usr/lib/spark/bin/compute-classpath.sh sets CLASSPATH="$SPARK_CLASSPATH". On CDH using parcels you can add the hive jars to SPARK_CLASSPATH like this:
SPARK_CLASSPATH=$(ls -1 /opt/cloudera/parcels/CDH/lib/hive/lib/*.jar | sed -e :a -e 'N;s/\n/:/;ta') /opt/cloudera/parcels/CDH/lib/spark/sbin/start-thriftserver.sh --help

Answer 4

Instructions from Cloudera Community forum
http://community.cloudera.com/t5/Advanced-Analytics-Apache-Spark/CDH-5-5-does-not-have-Spark-Thrift-Server/m-p/41849#M1758 :
git clone https://github.com/cloudera/spark.git
cd spark
./make-distribution.sh -DskipTests \
-Dhadoop.version=2.6.0-cdh5.7.0 \
-Phadoop-2.6 \
-Pyarn \
-Phive -Phive-thriftserver \
-Pflume-provided \
-Phadoop-provided \
-Phbase-provided \
-Phive-provided \
-Pparquet-provided
-Phive and -Phive-thriftserver are the key pieces there.
There is a request to add Spark Thrift Server
https://issues.cloudera.org/browse/DISTRO-817
please vote up if you want to see that in CDH.