Error mapping a nested format

[AndresUrregoAngel]

Hello guys,

After have load my DynamoDB table using scala when I try to retrieve a row filtered in a specific value I'm getting an error.

the full record format is :

{
  "agent_id": 14732,
  "called": "+185999999",
  "to_city": "null",
  "language": "english",
  "team_id": 15,
  "taskChannelUniqueName": "voice",
  "caller_country": "CA",
  "price": "-0.22880",
  "account_sid": "AC3bdb78f4397e3951251",
  "from": "+178049978523",
  "ivr_call_sid": "C983a7254cb0ade7862705",
  "travel_type": "international",
  "priceUnit": "USD",
  "called_zip": "null",
  "from_city": "EDMONTON",
  "caller_zip": "null",
  "from_state": "AB",
  "start_time": "2018-07-04 18:57:31",
  "site_id": "4",
  "status": "completed",
  "from_country": "CA",
  "original_task_sid": "XFT05514703a5c53ec0338a7ceb9cddf16c",
  "conference_sid": "CF6ffadad8e05be8b48d78c0704b21562c",
  "direct": "inbound",
  "to_country": "US",
  "conversations": {
    "handling_team_name": "Team 1",
    "conversation_attribute_1": "No",
    "hold_time": 12,
    "ivr_path": "other_inquiries",
    "ivr_time": 57,
    "conversation_id": "XDT55414703a5c53ec0338a7ceb9cddf16c",
    "segment": "CAe8e393775870c1ec4d765c2355009acd",
    "segment_link": [
      "https:\\/\\/d3dx7qtk6i17ic.cloudfront.net\\/production\\/2018\\/07\\/04\\/15\\/RE6eede8cf6d0a574a4f4564a32e555b38.wav"
    ],
    "external_contact": "JFLY TP",
    "case": 668132,
    "outcome": null,
    "queue": "Non-Revenue",
    "handling_team_id": 15,
    "in_business_hours": "Yes",
    "direction": "Inbound"
  },
  "to_state": "null",
  "products": {
    "product_attribute_1": "International",
    "brand": "JustFly",
    "product_attribute_2": "null"
  },
  "duration": "946",
  "booking_id": "108154082",
  "SourceApp": "ServPro",
  "call_sid": "CAb46c43e3e8a983a7254cb0ade7862705",
  "conversation_id": "XDT55414703a5c53ec0338a7ceb9cddf16c",
  "from_zip": "null",
  "customers": {
    "gender": "Male",
    "market_segment": "1st Time",
    "name": "Kofi Annor",
    "customer_id": 58484842,
    "business_value": "5596.19",
    "email": "[email protected]",
    "year_of_birth": 1521,
    "acquisition_date": "2017-12-15"
  },
  "department": "service",
  "external_contact": "JFLY TP",
  "call_type": "other_inquiries",
  "direction": "inbound",
  "support_case_id": 668132,
  "caller_state": "AB",
  "to_zip": "null",
  "called_country": "US",
  "end_time": "2018-07-04 19:13:17",
  "called_city": "null",
  "api_version": "2010-04-01",
  "called_state": "null",
  "agents": {
    "full_name": "Customer Name",
    "role": "agent_level_1",
    "agent_id": 14732,
    "manager": null,
    "location": null,
    "team_id": 15,
    "department": "service",
    "email": "[email protected]",
    "team_name": "Team 1"
  },
  "caller": "+19878874894",
  "caller_city": "EDMONTON",
  "to": "+15458779216"
}

Error:

at java.lang.Thread.run(Thread.java:748)
18/07/16 18:33:26 ERROR DynamoDBRelation: Failed converting item to row: {"conversation_id":"WT00014703a5c53ec0338a7ceb9cddf16c","agents":{"full_name":"Princess Naquita","role":"agent_level_1","agent_id":14732,"manager":"None","location":"None","team_id":15,"department":"service","email":"[email protected]","team_name":"Team 1"}}
scala.MatchError: StructType(StructField(agent_id,LongType,true), StructField(department,StringType,true), StructField(email,StringType,true), StructField(full_name,StringType,true), StructField(location,StringType,true), StructField(manager,StringType,true), StructField(role,StringType,true), StructField(team_id,LongType,true), StructField(team_name,StringType,true)) (of class org.apache.spark.sql.types.StructType)
at com.github.traviscrawford.spark.dynamodb.ItemConverter$$anonfun$1.apply(ItemConverter.scala:30)
at com.github.traviscrawford.spark.dynamodb.ItemConverter$$anonfun$1.apply(ItemConverter.scala:20)
at scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:234)
at scala.collection.TraversableLike$$anonfun$map$1.apply(TraversableLike.scala:234)
at scala.collection.Iterator$class.foreach(Iterator.scala:893)
at scala.collection.AbstractIterator.foreach(Iterator.scala:1336)
at scala.collection.IterableLike$class.foreach(IterableLike.scala:72)
at org.apache.spark.sql.types.StructType.foreach(StructType.scala:98)
at scala.collection.TraversableLike$class.map(TraversableLike.scala:234)
at org.apache.spark.sql.types.StructType.map(StructType.scala:98)
at com.github.traviscrawford.spark.dynamodb.ItemConverter$.toRow(ItemConverter.scala:20)
at com.github.traviscrawford.spark.dynamodb.DynamoDBRelation$$anonfun$scan$1$$anonfun$7.apply(DynamoDBRelation.scala:131)

[lockwobr]

was getting the same issue, and figured out how to get a round this... not ideal but seems to do the trick. Need to do more testing, but this might result in two reads, in which you might need to do add a persist or cach to the rdd so when it does the inferSchema its not read it again.

  import com.github.traviscrawford.spark.dynamodb._
  val tablename = "foobar"
  val region = "us-east-1"
  val totalSegments = 8
  val pageSize = 1000
  val table = spark.read.option("inferSchema", "true").json(DynamoScanner(spark.sparkContext, tablename, totalSegments, pageSize, None, None, Some(region)).toDS)

[AndresUrregoAngel]

I think this was solved by an incompability of my EMR I was running back on these days.