Spark 2.0 with Zeppelin 0.6.1 - SQLContext not available

Question

I am running spark 2.0 and zeppelin-0.6.1-bin-all on a Linux server. The default spark notebook runs just fine, but when I try to create and run a new notebook in pyspark using sqlContext I get the error "py4j.Py4JException: Method createDataFrame([class java.util.ArrayList, class java.util.ArrayList, null]) does not exist"

I tried running a simple code,

%pyspark

wordsDF = sqlContext.createDataFrame([('cat',), ('elephant',), ('rat',), ('rat',), ('cat', )], ['word'])
wordsDF.show()
print type(wordsDF)
wordsDF.printSchema()

I get the error,

Traceback (most recent call last):
  File "/tmp/zeppelin_pyspark-7635635698598314374.py", line 266, in 
    raise Exception(traceback.format_exc())
Exception: Traceback (most recent call last):
  File "/tmp/zeppelin_pyspark-7635635698598314374.py", line 259, in 
    exec(code)
  File "", line 1, in 
  File "/spark/spark-2.0.0-bin-hadoop2.7/python/pyspark/sql/context.py", line 299, in createDataFrame
    return self.sparkSession.createDataFrame(data, schema, samplingRatio)
  File "/spark/spark-2.0.0-bin-hadoop2.7/python/lib/py4j-0.10.1-src.zip/py4j/java_gateway.py", line 933, in __call__
    answer, self.gateway_client, self.target_id, self.name)
  File "/spark/spark-2.0.0-bin-hadoop2.7/python/pyspark/sql/utils.py", line 63, in deco
    return f(*a, **kw)
  File "/spark/spark-2.0.0-bin-hadoop2.7/python/lib/py4j-0.10.1-src.zip/py4j/protocol.py", line 316, in get_return_value
    format(target_id, ".", name, value))
Py4JError: An error occurred while calling o48.createDataFrame. Trace:
py4j.Py4JException: Method createDataFrame([class java.util.ArrayList, class java.util.ArrayList, null]) does not exist
    at py4j.reflection.ReflectionEngine.getMethod(ReflectionEngine.java:318)
    at py4j.reflection.ReflectionEngine.getMethod(ReflectionEngine.java:326)
    at py4j.Gateway.invoke(Gateway.java:272)
    at py4j.commands.AbstractCommand.invokeMethod(AbstractCommand.java:128)
    at py4j.commands.CallCommand.execute(CallCommand.java:79)
    at py4j.GatewayConnection.run(GatewayConnection.java:211)
    at java.lang.Thread.run(Thread.java:745)

When I try the same code with "sqlContext = SQLContext(sc)" it works just fine.

I have tried setting the interpreter "zeppelin.spark.useHiveContext false" configuration but it did not work.

I must obviously be missing something since this is such a simple operation. Please advice if there is any other configuration to be set or what I am missing.

I tested the same piece of code with Zeppelin 0.6.0 and it is working fine.

Alex Ioannides Alex Ioannides · Accepted Answer · 2016-09-29T11:39:35

SparkSession is the default entry-point for Spark 2.0.0, which is mapped to spark in Zeppelin 0.6.1 (as it is in the Spark shell). Have you tried spark.createDataFrame(...)?

Spark 2.0 with Zeppelin 0.6.1 - SQLContext not available

1 Answers