I am running CDH 4.4 with Spark 0.9.0 from a Cloudera parcel.
I have a bunch of Avro files that were created via Pig's AvroStorage UDF. I want to load these files in Spark, using a generic record or the schema onboard the Avro files. So far I've tried this:
import org.apache.avro.mapred.AvroKey
import org.apache.avro.mapreduce.AvroKeyInputFormat
import org.apache.hadoop.io.NullWritable
import org.apache.commons.lang.StringEscapeUtils.escapeCsv
import org.apache.hadoop.fs.Path
import org.apache.hadoop.fs.FileSystem
import org.apache.hadoop.conf.Configuration
import java.net.URI
import java.io.BufferedInputStream
import java.io.File
import org.apache.avro.generic.{GenericDatumReader, GenericRecord}
import org.apache.avro.specific.SpecificDatumReader
import org.apache.avro.file.DataFileStream
import org.apache.avro.io.DatumReader
import org.apache.avro.file.DataFileReader
import org.apache.avro.mapred.FsInput
val input = "hdfs://hivecluster2/securityx/web_proxy_mef/2014/05/29/22/part-m-00016.avro"
val inURI = new URI(input)
val inPath = new Path(inURI)
val fsInput = new FsInput(inPath, sc.hadoopConfiguration)
val reader = new GenericDatumReader[GenericRecord]
val dataFileReader = DataFileReader.openReader(fsInput, reader)
val schemaString = dataFileReader.getSchema
val buf = scala.collection.mutable.ListBuffer.empty[GenericRecord]
while(dataFileReader.hasNext) {
buf += dataFileReader.next
}
sc.parallelize(buf)
This works for one file, but it can't scale - I am loading all the data into local RAM and then distributing it across the spark nodes from there.
To answer my own question:
import org.apache.spark.SparkContext
import org.apache.spark.SparkContext._
import org.apache.avro.generic.GenericRecord
import org.apache.avro.mapred.AvroKey
import org.apache.avro.mapred.AvroInputFormat
import org.apache.avro.mapreduce.AvroKeyInputFormat
import org.apache.hadoop.io.NullWritable
import org.apache.commons.lang.StringEscapeUtils.escapeCsv
import org.apache.hadoop.fs.FileSystem
import org.apache.hadoop.fs.Path
import org.apache.hadoop.conf.Configuration
import java.io.BufferedInputStream
import org.apache.avro.file.DataFileStream
import org.apache.avro.io.DatumReader
import org.apache.avro.file.DataFileReader
import org.apache.avro.file.DataFileReader
import org.apache.avro.generic.{GenericDatumReader, GenericRecord}
import org.apache.avro.mapred.FsInput
import org.apache.avro.Schema
import org.apache.avro.Schema.Parser
import org.apache.hadoop.mapred.JobConf
import java.io.File
import java.net.URI
// spark-shell -usejavacp -classpath "*.jar"
val input = "hdfs://hivecluster2/securityx/web_proxy_mef/2014/05/29/22/part-m-00016.avro"
val jobConf= new JobConf(sc.hadoopConfiguration)
val rdd = sc.hadoopFile(
input,
classOf[org.apache.avro.mapred.AvroInputFormat[GenericRecord]],
classOf[org.apache.avro.mapred.AvroWrapper[GenericRecord]],
classOf[org.apache.hadoop.io.NullWritable],
10)
val f1 = rdd.first
val a = f1._1.datum
a.get("rawLog") // Access avro fields
This works for me:
import org.apache.avro.generic.GenericRecord
import org.apache.avro.mapred.{AvroInputFormat, AvroWrapper}
import org.apache.hadoop.io.NullWritable
...
val path = "hdfs:///path/to/your/avro/folder"
val avroRDD = sc.hadoopFile[AvroWrapper[GenericRecord], NullWritable, AvroInputFormat[GenericRecord]](path)
Related
I am facing an issue during spark streaming. I am getting empty records after it gets streamed and passes to the "parse" method.
My code:
import spark.implicits._
import org.apache.spark.sql.types._
import org.apache.spark.sql.Encoders
import org.apache.spark.streaming._
import org.apache.spark.sql.functions._
import org.apache.spark.sql.SparkSession
import spark.implicits._
import org.apache.spark.sql.types.{StructType, StructField, StringType,
IntegerType}
import org.apache.spark.sql.functions._
import org.apache.spark.sql.SparkSession
import spark.implicits._
import org.apache.spark.sql.types.{StructType, StructField, StringType,
IntegerType}
import org.apache.spark.SparkConf
import org.apache.spark.streaming.{Seconds, StreamingContext}
import org.apache.spark.storage.StorageLevel
import java.util.regex.Pattern
import java.util.regex.Matcher
import org.apache.spark.sql.hive.HiveContext;
import org.apache.spark.sql.streaming.Trigger
import org.apache.spark.sql._
val conf = new SparkConf().setAppName("streamHive").setMaster("local[*]").set("spark.driver.allowMultipleContexts", "true")
val ssc = new StreamingContext(conf, Seconds(5))
val sc=ssc.sparkContext
val lines = ssc.textFileStream("file:///home/sadr/testHive")
case class Prices(name: String, age: String,sex: String, location: String)
val sqlContext = new org.apache.spark.sql.SQLContext(sc)
def parse (rdd : org.apache.spark.rdd.RDD[String] ) =
{
var l = rdd.map(_.split(","))
val prices = l.map(p => Prices(p(0),p(1),p(2),p(3)))
val pricesDf = sqlContext.createDataFrame(prices)
pricesDf.registerTempTable("prices")
pricesDf.show()
var x = sqlContext.sql("select count(*) from prices")
x.show()}
lines.foreachRDD { rdd => parse(rdd)}
lines.print()
ssc.start()
My input file:
cat test1.csv
Riaz,32,M,uk
tony,23,M,india
manu,33,M,china
imart,34,F,AUS
I am getting this output:
lines.foreachRDD { rdd => parse(rdd)}
lines.print()
ssc.start()
scala> +----+---+---+--------+
|name|age|sex|location|
+----+---+---+--------+
+----+---+---+--------+
I am using Spark version 2.3....I AM GETTING FOLLOWING ERROR AFTER ADDING X.SHOW()
Not sure if you are actually able to read the streams.
textFileStream reads only the new files added to the directory after the program starts and not the existing ones. Was the file already there?
If yes, remove it from the directory, start the program and copy the file again?
hello i just started to learn scala.
and just follow the tutorial in udemy.
i was followed the same code but give me an error.
i have no idea about that error.
and this my code
import org.apache.spark.ml.classification.LogisticRegression
import org.apache.spark.sql.SparkSession
import org.apache.log4j._
import org.apache.spark.ml.feature.{CountVectorizer, CountVectorizerModel}
import org.apache.spark.ml.feature.Word2Vec
import org.apache.spark.ml.linalg.Vector
import org.apache.spark.sql.Row
Logger.getLogger("org").setLevel(Level.ERROR)
val spark = SparkSession.builder().getOrCreate()
val data = spark.read.option("header","true").
option("inferSchema","true").
option("delimiter","\t").
format("csv").
load("dataset.tsv").
withColumn("subject", split($"subject", " "))
val logRegDataAll = (data.select(data("label")).as("label"),$"subject")
val logRegData = logRegDataAll.na.drop()
and give me error like this
scala> :load LogisticRegression.scala
Loading LogisticRegression.scala...
import org.apache.spark.ml.classification.LogisticRegression
import org.apache.spark.sql.SparkSession
import org.apache.log4j._
import org.apache.spark.ml.feature.{CountVectorizer, CountVectorizerModel}
import org.apache.spark.ml.feature.Word2Vec
import org.apache.spark.ml.linalg.Vector
import org.apache.spark.sql.Row
spark: org.apache.spark.sql.SparkSession = org.apache.spark.sql.SparkSession#1efcba00
data: org.apache.spark.sql.DataFrame = [label: string, subject: array<string>]
logRegDataAll: (org.apache.spark.sql.Dataset[org.apache.spark.sql.Row], org.apache.spark.sql.ColumnName) = ([label: string],subject)
<console>:43: error: value na is not a member of (org.apache.spark.sql.Dataset[org.apache.spark.sql.Row], org.apache.spark.sql.ColumnName)
val logRegData = logRegDataAll.na.drop()
^
thanks for helping
You can see clearly
val logRegDataAll = (data.select(data("label")).as("label"),$"subject")
This returns
(org.apache.spark.sql.Dataset[org.apache.spark.sql.Row], org.apache.spark.sql.ColumnName)
So there is an extra parantheses ) data("label")) which should be data.select(data("label").as("label"),$"subject") in actual.
I am trying to parse data from a XML file through Spark using databrics library
Here is my code:
import org.apache.spark.SparkConf
import org.apache.spark.SparkContext
import org.apache.spark.sql.SQLContext
import org.apache.spark.sql.functions
import java.text.Format
import org.apache.spark.sql.functions.concat_ws
import org.apache.spark.sql
import org.apache.spark.sql.types._
import org.apache.spark.sql.catalyst.plans.logical.With
import org.apache.spark.sql.functions.lit
import org.apache.spark.sql.functions.udf
import scala.sys.process._
import org.apache.spark.sql.functions.lit
import org.apache.spark.sql.functions.udf
import org.apache.spark.sql.functions._
object printschema
{
def main(args: Array[String]): Unit =
{
val conf = new SparkConf().setAppName("printschema").setMaster("local")
conf.set("spark.debug.maxToStringFields", "10000000")
val context = new SparkContext(conf)
val sqlCotext = new SQLContext(context)
import sqlCotext.implicits._
val df = sqlCotext.read.format("com.databricks.spark.xml")
.option("rowTag", "us-bibliographic-data-application")
.option("treatEmptyValuesAsNulls", true)
.load("/Users/praveen/Desktop/ipa0105.xml")
val q1= df.withColumn("document",$"application-reference.document-id.doc-number".cast(sql.types.StringType))
.withColumn("document_number",$"application-reference.document-id.doc-number".cast(sql.types.StringType)).select("document","document_number").collect()
for(l<-q1)
{
val m1=l.get(0)
val m2=l.get(1)
println(m1,m2)
}
}
}
When I run the code on ScalaIDE/IntelliJ IDEA it works fine and here is my Output.
(14789882,14789882)
(14755945,14755945)
(14755919,14755919)
But, when I build a jar and execute it by using spark-submit it returns simply null values
OUTPUT :
NULL,NULL
NULL,NULL
NULL,NULL
Here is my Spark submit:
./spark-submit --jars /home/hadoop/spark-xml_2.11-0.4.0.jar --class inndata.praveen --master local[2] /home/hadoop/ip/target/scala-2.11/ip_2.11-1.0.jar
I am very much new to Scala and need to write a test case which will mock the Postgresql connections and Statements.However unable to do so and getting the error.Can anyone help me.Below is the code that I've written
Thanks in advance !!
import org.apache.spark.sql.types.{StringType, StructField, StructType}
import org.apache.spark.sql.Column`
import org.slf4j.LoggerFactory
import java.nio.file.Paths
import java.sql.ResultSet
import java.io.InputStream
import java.io.Reader
import java.util
import java.io.File
import java.util.UUID
import java.nio.file.attribute.PosixFilePermission
import com.typesafe.config.ConfigFactory
import org.apache.spark.sql.{DataFrame, SQLContext}
import org.scalatest.{Matchers, WordSpecLike, BeforeAndAfter}
import org.scalactic.{Good, Bad, Many, One}
import scala.collection.JavaConverters._
import spark.jobserver.{SparkJobValid, SparkJobInvalid}
import spark.jobserver.api.{JobEnvironment, SingleProblem}
import org.apache.spark.sql.{Column, Row, DataFrame}
import java.sql.Connection
import java.sql.DriverManager
import java.sql.ResultSet
import org.junit.Assert
import org.junit.Before
import org.junit.Test
import org.junit.runner.RunWith
import org.easymock.EasyMock.expect
import org.powermock.api._
import org.powermock.core.classloader.annotations.PrepareForTest
import java.io.FileReader
import org.scalamock.scalatest.MockFactory
import org.powermock.core.classloader.annotations.PrepareForTest
import org.powermock.api.mockito.PowerMockito
import org.powermock.api.mockito.PowerMockito._
import org.postgresql.copy.CopyManager
import scala.collection.JavaConversions._
import org.mockito.Matchers.any
import java.sql.Statement
class mockCopyManager(){
def copyIn(command : String , fR:java.io.FileReader) :Unit ={
println("Run Command {}".format(command))
}
}
class AdvisoretlSpec extends WordSpecLike with Matchers with
MockFactory {
val sc = SparkUnitTestContext.hiveContext
import SparkUnitTestContext.defaultSizeInBytes
"Class Advisoretl job" should {
"load data in "{
val csvMap : Map[String,String] = Map("t1"->"t1.csv","t2"->"t2.csv")
val testObj = new Advisoretl()
val mockStatement = mock[Statement]
val mockConnection=mock[Connection]
val a:String = "TRUNCATE TABLE t1"
val b:String = "TRUNCATE TABLE t2"
PowerMockito.mockStatic(classOf[DriverManager])
val mockCopyManager=mock[CopyManager]
PowerMockito.when(DriverManager.getConnection(any[String]), Nil: _*).thenReturn(mockConnection)
(mockConnection.createStatement _).when().returns(mockStatement)
(mockStatement.executeUpdate _).when(a).returns(1)
(mockStatement.executeUpdate _).when("TRUNCATE TABLE t2").returns(1)
(mockCopyManager.copyIn _).when(*).returns(1)*/
val fnResult = testObj.connectionWithPostgres("a", "b", "c", "target/testdata", csvMap)
fnResult should be ("OK")
}
}
}'
Issue Description:
Spark Version: 1.6.2
Execution: Spark-shell (REPL) master = local[2] (tried local[*])
example.json is as below:
{"name":"D2" ,"lovesPandas":"Y"}
{"name":"D3" ,"lovesPandas":"Y"}
{"name":"D4" ,"lovesPandas":"Y"}
{"name":"D5" ,"lovesPandas":"Y"}
Code executing in Spark-shell local mode:
import org.apache.spark._
import org.apache.spark.streaming._
import org.apache.spark.streaming.StreamingContext._
import org.apache.spark.streaming.kafka._
import org.apache.spark.sql._
import org.json4s._
import org.json4s.jackson.JsonMethods._
import _root_.kafka.serializer.StringDecoder
import _root_.kafka.serializer.Decoder
import _root_.kafka.utils.VerifiableProperties
import org.apache.hadoop.hbase._
import org.apache.hadoop.hbase.HBaseConfiguration
import org.apache.hadoop.hbase.client.Put
import org.apache.hadoop.hbase.io.ImmutableBytesWritable
import org.apache.hadoop.hbase.mapred.TableOutputFormat
import org.apache.hadoop.hbase.util.Bytes
import org.apache.hadoop.mapred.JobConf
import org.apache.kafka.clients.consumer.ConsumerConfig
import org.apache.spark.SparkConf
import org.apache.spark.SparkContext
import org.apache.spark.streaming.Seconds
import org.apache.spark.streaming.StreamingContext
val ssc = new StreamingContext(sc, Seconds(2) )
val messages = ssc.textFileStream("C:\\pdtemp\\test\\example.json")
messages.print()
I tried the saveAsTextFiles but it is not saving any files too.
this does not work -- shows no output -- tried the same with reading stream from Kafka on a spark-shell
tried the following too -- does not work:
messages.foreachRDD(rdd => rdd.foreach(print))
Also, tried parsing the schema converting to dataframe but nothing seems to work
normal json parsing is working and i can print the contents of normal //RDD/DF //to console in Spark-shell
Can anyone help, please?