Hi, We have a shared CDH 5.3.3 cluster and trying to use Spark 1.5.1 on it in yarn client mode with Hive.
I have compiled Spark 1.5.1 with SPARK_HIVE=true, but it seems I am not able to make SparkSQL to pick up the hive-site.xml when runnig pyspark. hive-site.xml is located in $SPARK_HOME/hadoop-conf/hive-site.xml and also in $SPARK_HOME/conf/hive-site.xml When I start pyspark with the below command and then run some simple SparkSQL it fails, it seems it didn't pic up the settings in hive-site.xml $ HADOOP_CONF_DIR=$SPARK_HOME/hadoop-conf YARN_CONF_DIR=$SPARK_HOME/yarn-conf HADOOP_USER_NAME=biapp MASTER=yarn $SPARK_HOME/bin/pyspark --deploy-mode client Python 2.6.6 (r266:84292, Jul 23 2015, 05:13:40) [GCC 4.4.7 20120313 (Red Hat 4.4.7-16)] on linux2 Type "help", "copyright", "credits" or "license" for more information. SLF4J: Class path contains multiple SLF4J bindings. SLF4J: Found binding in [jar:file:/usr/lib/spark-1.5.1-bin-without-hadoop/lib/spark-assembly-1.5.1-hadoop2.5.0-cdh5.3.3.jar!/org/slf4j/impl/StaticLoggerBinder.class] SLF4J: Found binding in [jar:file:/usr/lib/zookeeper/lib/slf4j-log4j12-1.7.5.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] 15/10/28 10:22:33 WARN MetricsSystem: Using default name DAGScheduler for source because spark.app.id is not set. 15/10/28 10:22:35 WARN NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable 15/10/28 10:22:59 WARN HiveConf: HiveConf of name hive.metastore.local does not exist Welcome to ____ __ / __/__ ___ _____/ /__ _\ \/ _ \/ _ `/ __/ '_/ /__ / .__/\_,_/_/ /_/\_\ version 1.5.1 /_/ Using Python version 2.6.6 (r266:84292, Jul 23 2015 05:13:40) SparkContext available as sc, HiveContext available as sqlContext. >>> sqlContext2 = HiveContext(sc) >>> sqlContext2.sql("show databases").first() 15/10/28 10:23:12 WARN HiveConf: HiveConf of name hive.metastore.local does not exist 15/10/28 10:23:13 WARN ShellBasedUnixGroupsMapping: got exception trying to get groups for user biapp: id: biapp: No such user 15/10/28 10:23:13 WARN UserGroupInformation: No groups available for user biapp Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/usr/lib/spark-1.5.1-bin-without-hadoop/python/pyspark/sql/context.py", line 552, in sql return DataFrame(self._ssql_ctx.sql(sqlQuery), self) File "/usr/lib/spark-1.5.1-bin-without-hadoop/python/pyspark/sql/context.py", line 660, in _ssql_ctx "build/sbt assembly", e) Exception: ("You must build Spark with Hive. Export 'SPARK_HIVE=true' and run build/sbt assembly", Py4JJavaError(u'An error occurred while calling None.org.apache.spark.sql.hive.HiveContext.\n', JavaObject id=o20)) >>> See in the above the warning about "WARN HiveConf: HiveConf of name hive.metastore.local does not exist" while actually there is a hive.metastore.local attribute in the hive-site.xml Any idea how to submit hive-site.xml in yarn client mode? Thanks