Recently I discovered how to load Avros in Spark using a Hadoop InputFormat.

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