SiddharthRam created SPARK-20488:
------------------------------------
Summary: how to create a dataframe or dataset in a streaming
context spark 2.1.0 on mesos 1.0,1
Key: SPARK-20488
URL: https://issues.apache.org/jira/browse/SPARK-20488
Project: Spark
Issue Type: Question
Components: Spark Submit
Affects Versions: 2.1.0
Reporter: SiddharthRam
Please suggest the right way
below is one attempt from my end :
i`m using spark 2.1.0 (spark-submit job) and below is a snippet of the code i`m
using and always getting a null pointer exception when trying to create a
dataframe or dataset
val sparkConf = new SparkConf().setAppName(Kafka_topic)
val warehouseLocation = "file:${system:user.dir}/spark-warehouse"
val spark = SparkSession
.builder
.config(sparkConf)
.config("spark.sql.warehouse.dir", warehouseLocation)
.enableHiveSupport()
.getOrCreate()
val ssc = new StreamingContext(spark.sparkContext,Seconds(1))
val stream = KafkaUtils.createDirectStream[String, String](
ssc,
PreferConsistent,
Subscribe[String, String](topics, kafkaParams)
stream.foreachRDD( rdd => {
if(!rdd.isEmpty()) {
val rdm = rdd.map(rdd =>
(rdd.key(),rdd.value(),rdd.partition(),rdd.offset()))
rdm.foreach(rda => {
// bunch of code works fine
//load json from file or String
var sqlContext = spark.read.json((compact(render(json)))) [ load example ]
}
})
}
ssc.start()
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
