So I have been banging my head against this for the last couple of days. I am having trouble de-serializing an Avro file that we are generating and sending into Azure Event Hub. We are attempting to do this with Databricks Runtime 7.2 Structured Streaming. Using the newer from_avro method described here to de-serialize the body of the event message.
import org.apache.spark.eventhubs._
import org.apache.spark.sql.functions._
import org.apache.spark.sql.avro._
import org.apache.avro._
import org.apache.spark.sql.types._
import org.apache.spark.sql.avro.functions._
val connStr = "<EventHubConnectionstring>"
val customEventhubParameters =
EventHubsConf(connStr.toString())
.setMaxEventsPerTrigger(5)
//.setStartingPosition(EventPosition.fromStartOfStream)
val incomingStream = spark
.readStream
.format("eventhubs")
.options(customEventhubParameters.toMap)
.load()
.filter($"properties".getItem("TableName") === "Branches")
val avroSchema = s"""{"type":"record","name":"Branches","fields":[{"name":"_src_ChangeOperation","type":["null","string"]},{"name":"_src_CurrentTrackingId","type":["null","long"]},{"name":"_src_RecordExtractUTCTimestamp","type":"string"},{"name":"ID","type":["null","int"]},{"name":"BranchCode","type":["null","string"]},{"name":"BranchName","type":["null","string"]},{"name":"Address1","type":["null","string"]},{"name":"Address2","type":["null","string"]},{"name":"City","type":["null","string"]},{"name":"StateID","type":["null","int"]},{"name":"ZipCode","type":["null","string"]},{"name":"Telephone","type":["null","string"]},{"name":"Contact","type":["null","string"]},{"name":"Title","type":["null","string"]},{"name":"DOB","type":["null","string"]},{"name":"TimeZoneID","type":["null","int"]},{"name":"ObserveDaylightSaving","type":["null","boolean"]},{"name":"PaySummerTimeHour","type":["null","boolean"]},{"name":"PayWinterTimeHour","type":["null","boolean"]},{"name":"BillSummerTimeHour","type":["null","boolean"]},{"name":"BillWinterTimeHour","type":["null","boolean"]},{"name":"Deleted","type":["null","boolean"]},{"name":"LastUpdated","type":["null","string"]},{"name":"txJobID","type":["null","string"]},{"name":"SourceID","type":["null","string"]},{"name":"HP_UseHolPayHourMethod","type":["null","boolean"]},{"name":"HP_HourlyRatePercent","type":["null","float"]},{"name":"HP_RequiredWeeksOfEmployment","type":["null","float"]},{"name":"rgUseSystemSettings","type":["null","boolean"]},{"name":"rgDutySplitBy","type":["null","int"]},{"name":"rgBasePeriodDate","type":["null","string"]},{"name":"rgFirstDayOfWeek","type":["null","int"]},{"name":"rgDutyStartOfDayTime","type":["null","string"]},{"name":"rgHolidayStartOfDayTime","type":["null","string"]},{"name":"rgMinimumTimePeriod","type":["null","int"]},{"name":"rgLoadPublicTable","type":["null","boolean"]},{"name":"rgPOTPayPeriodID","type":["null","int"]},{"name":"rgPOT1","type":["null","string"]},{"name":"rgPOT2","type":["null","string"]},{"name":"Facsimile","type":["null","string"]},{"name":"CountryID","type":["null","int"]},{"name":"EmailAddress","type":["null","string"]},{"name":"ContractSecurityHistoricalWeeks","type":["null","int"]},{"name":"ContractSecurityFutureWeeks","type":["null","int"]},{"name":"TimeLinkTelephone1","type":["null","string"]},{"name":"TimeLinkTelephone2","type":["null","string"]},{"name":"TimeLinkTelephone3","type":["null","string"]},{"name":"TimeLinkTelephone4","type":["null","string"]},{"name":"TimeLinkTelephone5","type":["null","string"]},{"name":"AutoTakeMissedCalls","type":["null","boolean"]},{"name":"AutoTakeMissedCallsDuration","type":["null","string"]},{"name":"AutoTakeApplyDurationToCheckCalls","type":["null","boolean"]},{"name":"AutoTakeMissedCheckCalls","type":["null","boolean"]},{"name":"AutoTakeMissedCheckCallsDuration","type":["null","string"]},{"name":"DocumentLocation","type":["null","string"]},{"name":"DefaultPortalAccess","type":["null","boolean"]},{"name":"DefaultPortalSecurityRoleID","type":["null","int"]},{"name":"EmployeeTemplateID","type":["null","int"]},{"name":"SiteCardTemplateID","type":["null","int"]},{"name":"TSAllowancesHeaderID","type":["null","int"]},{"name":"TSMinimumWageHeaderID","type":["null","int"]},{"name":"TimeLinkClaimMade","type":["null","boolean"]},{"name":"TSAllowancePeriodBaseDate","type":["null","string"]},{"name":"TSAllowancePeriodID","type":["null","int"]},{"name":"TSMinimumWageCalcMethodID","type":["null","int"]},{"name":"FlexibleShiftsHeaderID","type":["null","int"]},{"name":"SchedulingUseSystemSettings","type":["null","boolean"]},{"name":"MinimumRestPeriod","type":["null","int"]},{"name":"TSMealBreakHeaderID","type":["null","int"]},{"name":"ServiceTracImportType","type":["null","int"]},{"name":"StandDownDiaryEventID","type":["null","int"]},{"name":"ScheduledDutyChangeMessageTemplateId","type":["null","int"]},{"name":"ScheduledDutyAddedMessageTemplateId","type":["null","int"]},{"name":"ScheduledDutyRemovedMessageTemplateId","type":["null","int"]},{"name":"NegativeMessageResponsesPermitted","type":["null","boolean"]},{"name":"PortalEventsStandardLocFirst","type":["null","boolean"]},{"name":"ReminderMessage","type":["null","boolean"]},{"name":"ReminderMessageDaysBefore","type":["null","int"]},{"name":"ReminderMessageTemplateId","type":["null","int"]},{"name":"ScheduledDutyChangeMessageAllowReply","type":["null","boolean"]},{"name":"ScheduledDutyAddedMessageAllowReply","type":["null","boolean"]},{"name":"PayAlertEscalationGroup","type":["null","int"]},{"name":"BudgetedPay","type":["null","int"]},{"name":"PayAlertVariance","type":["null","string"]},{"name":"BusinessUnitID","type":["null","int"]},{"name":"APH_Hours","type":["null","float"]},{"name":"APH_Period","type":["null","int"]},{"name":"APH_PeriodCount","type":["null","int"]},{"name":"AveragePeriodHoursRuleId","type":["null","int"]},{"name":"HolidayScheduleID","type":["null","int"]},{"name":"AutomationRuleProfileId","type":["null","int"]}]}"""
val decoded_df = incomingStream
.select(
from_avro($"body",avroSchema).alias("payload")
)
val query1 = (
decoded_df
.writeStream
.format("memory")
.queryName("read_hub")
.start()
)
I have verified that the file we are sending has a valid schema, that it has data in it and that it is getting to the stream job in the notebook before failing with the following stack trace that states that the data is malformed. However I am able to write the generated file to a .avro file and de-serialize it using the normal .read.format("avro") method just fine.
at org.apache.spark.sql.execution.datasources.v2.V2TableWriteExec.writeWithV2(WriteToDataSourceV2Exec.scala:413)
at org.apache.spark.sql.execution.datasources.v2.V2TableWriteExec.writeWithV2$(WriteToDataSourceV2Exec.scala:361)
at org.apache.spark.sql.execution.datasources.v2.WriteToDataSourceV2Exec.writeWithV2(WriteToDataSourceV2Exec.scala:322)
at org.apache.spark.sql.execution.datasources.v2.WriteToDataSourceV2Exec.run(WriteToDataSourceV2Exec.scala:329)
at org.apache.spark.sql.execution.datasources.v2.V2CommandExec.result$lzycompute(V2CommandExec.scala:39)
at org.apache.spark.sql.execution.datasources.v2.V2CommandExec.result(V2CommandExec.scala:39)
at org.apache.spark.sql.execution.datasources.v2.V2CommandExec.executeCollect(V2CommandExec.scala:45)
at org.apache.spark.sql.execution.collect.Collector$.callExecuteCollect(Collector.scala:118)
at org.apache.spark.sql.execution.collect.Collector$.collect(Collector.scala:69)
at org.apache.spark.sql.execution.collect.Collector$.collect(Collector.scala:88)
at org.apache.spark.sql.execution.ResultCacheManager.getOrComputeResult(ResultCacheManager.scala:508)
at org.apache.spark.sql.execution.ResultCacheManager.getOrComputeResult(ResultCacheManager.scala:480)
at org.apache.spark.sql.execution.SparkPlan.executeCollectResult(SparkPlan.scala:396)
at org.apache.spark.sql.Dataset.collectResult(Dataset.scala:2986)
at org.apache.spark.sql.Dataset.collectFromPlan(Dataset.scala:3692)
at org.apache.spark.sql.Dataset.$anonfun$collect$1(Dataset.scala:2953)
at org.apache.spark.sql.Dataset.$anonfun$withAction$1(Dataset.scala:3684)
at org.apache.spark.sql.execution.SQLExecution$.$anonfun$withCustomExecutionEnv$5(SQLExecution.scala:116)
at org.apache.spark.sql.execution.SQLExecution$.withSQLConfPropagated(SQLExecution.scala:248)
at org.apache.spark.sql.execution.SQLExecution$.$anonfun$withCustomExecutionEnv$1(SQLExecution.scala:101)
at org.apache.spark.sql.SparkSession.withActive(SparkSession.scala:835)
at org.apache.spark.sql.execution.SQLExecution$.withCustomExecutionEnv(SQLExecution.scala:77)
at org.apache.spark.sql.execution.SQLExecution$.withNewExecutionId(SQLExecution.scala:198)
at org.apache.spark.sql.Dataset.withAction(Dataset.scala:3682)
at org.apache.spark.sql.Dataset.collect(Dataset.scala:2953)
at org.apache.spark.sql.execution.streaming.MicroBatchExecution.$anonfun$runBatch$16(MicroBatchExecution.scala:586)
at org.apache.spark.sql.execution.SQLExecution$.$anonfun$withCustomExecutionEnv$5(SQLExecution.scala:116)
at org.apache.spark.sql.execution.SQLExecution$.withSQLConfPropagated(SQLExecution.scala:248)
at org.apache.spark.sql.execution.SQLExecution$.$anonfun$withCustomExecutionEnv$1(SQLExecution.scala:101)
at org.apache.spark.sql.SparkSession.withActive(SparkSession.scala:835)
at org.apache.spark.sql.execution.SQLExecution$.withCustomExecutionEnv(SQLExecution.scala:77)
at org.apache.spark.sql.execution.SQLExecution$.withNewExecutionId(SQLExecution.scala:198)
at org.apache.spark.sql.execution.streaming.MicroBatchExecution.$anonfun$runBatch$15(MicroBatchExecution.scala:581)
at org.apache.spark.sql.execution.streaming.ProgressReporter.reportTimeTaken(ProgressReporter.scala:276)
at org.apache.spark.sql.execution.streaming.ProgressReporter.reportTimeTaken$(ProgressReporter.scala:274)
at org.apache.spark.sql.execution.streaming.StreamExecution.reportTimeTaken(StreamExecution.scala:71)
at org.apache.spark.sql.execution.streaming.MicroBatchExecution.runBatch(MicroBatchExecution.scala:581)
at org.apache.spark.sql.execution.streaming.MicroBatchExecution.$anonfun$runActivatedStream$2(MicroBatchExecution.scala:231)
at scala.runtime.java8.JFunction0$mcV$sp.apply(JFunction0$mcV$sp.java:23)
at org.apache.spark.sql.execution.streaming.ProgressReporter.reportTimeTaken(ProgressReporter.scala:276)
at org.apache.spark.sql.execution.streaming.ProgressReporter.reportTimeTaken$(ProgressReporter.scala:274)
at org.apache.spark.sql.execution.streaming.StreamExecution.reportTimeTaken(StreamExecution.scala:71)
at org.apache.spark.sql.execution.streaming.MicroBatchExecution.$anonfun$runActivatedStream$1(MicroBatchExecution.scala:199)
at org.apache.spark.sql.execution.streaming.ProcessingTimeExecutor.execute(TriggerExecutor.scala:57)
at org.apache.spark.sql.execution.streaming.MicroBatchExecution.runActivatedStream(MicroBatchExecution.scala:193)
at org.apache.spark.sql.execution.streaming.StreamExecution.org$apache$spark$sql$execution$streaming$StreamExecution$$runStream(StreamExecution.scala:346)
at org.apache.spark.sql.execution.streaming.StreamExecution$$anon$1.run(StreamExecution.scala:259)
Caused by: org.apache.spark.SparkException: Job aborted due to stage failure: Task 0 in stage 37.0 failed 4 times, most recent failure: Lost task 0.3 in stage 37.0 (TID 84, 10.139.64.5, executor 0): org.apache.spark.SparkException: Malformed records are detected in record parsing. Current parse Mode: FAILFAST. To process malformed records as null result, try setting the option 'mode' as 'PERMISSIVE'.
at org.apache.spark.sql.avro.AvroDataToCatalyst.nullSafeEval(AvroDataToCatalyst.scala:111)
at org.apache.spark.sql.catalyst.expressions.GeneratedClass$GeneratedIteratorForCodegenStage1.processNext(Unknown Source)
at org.apache.spark.sql.execution.BufferedRowIterator.hasNext(BufferedRowIterator.java:43)
at org.apache.spark.sql.execution.WholeStageCodegenExec$$anon$1.hasNext(WholeStageCodegenExec.scala:731)
at org.apache.spark.sql.execution.datasources.v2.DataWritingSparkTask$.$anonfun$run$7(WriteToDataSourceV2Exec.scala:438)
at org.apache.spark.util.Utils$.tryWithSafeFinallyAndFailureCallbacks(Utils.scala:1615)
at org.apache.spark.sql.execution.datasources.v2.DataWritingSparkTask$.run(WriteToDataSourceV2Exec.scala:477)
at org.apache.spark.sql.execution.datasources.v2.V2TableWriteExec.$anonfun$writeWithV2$2(WriteToDataSourceV2Exec.scala:385)
at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:90)
at org.apache.spark.scheduler.Task.doRunTask(Task.scala:144)
at org.apache.spark.scheduler.Task.run(Task.scala:117)
at org.apache.spark.executor.Executor$TaskRunner.$anonfun$run$9(Executor.scala:657)
at org.apache.spark.util.Utils$.tryWithSafeFinally(Utils.scala:1581)
at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:660)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
Caused by: java.lang.ArrayIndexOutOfBoundsException: -40
at org.apache.avro.io.parsing.Symbol$Alternative.getSymbol(Symbol.java:424)
at org.apache.avro.io.ResolvingDecoder.doAction(ResolvingDecoder.java:290)
at org.apache.avro.io.parsing.Parser.advance(Parser.java:88)
at org.apache.avro.io.ResolvingDecoder.readIndex(ResolvingDecoder.java:267)
at org.apache.avro.generic.GenericDatumReader.readWithoutConversion(GenericDatumReader.java:179)
at org.apache.avro.generic.GenericDatumReader.read(GenericDatumReader.java:153)
at org.apache.avro.generic.GenericDatumReader.readField(GenericDatumReader.java:232)
at org.apache.avro.generic.GenericDatumReader.readRecord(GenericDatumReader.java:222)
at org.apache.avro.generic.GenericDatumReader.readWithoutConversion(GenericDatumReader.java:175)
at org.apache.avro.generic.GenericDatumReader.read(GenericDatumReader.java:153)
at org.apache.avro.generic.GenericDatumReader.read(GenericDatumReader.java:145)
at org.apache.spark.sql.avro.AvroDataToCatalyst.nullSafeEval(AvroDataToCatalyst.scala:100)
... 16 more
Driver stacktrace:
at org.apache.spark.scheduler.DAGScheduler.failJobAndIndependentStages(DAGScheduler.scala:2478)
at org.apache.spark.scheduler.DAGScheduler.$anonfun$abortStage$2(DAGScheduler.scala:2427)
at org.apache.spark.scheduler.DAGScheduler.$anonfun$abortStage$2$adapted(DAGScheduler.scala:2426)
at scala.collection.mutable.ResizableArray.foreach(ResizableArray.scala:62)
at scala.collection.mutable.ResizableArray.foreach$(ResizableArray.scala:55)
at scala.collection.mutable.ArrayBuffer.foreach(ArrayBuffer.scala:49)
at org.apache.spark.scheduler.DAGScheduler.abortStage(DAGScheduler.scala:2426)
at org.apache.spark.scheduler.DAGScheduler.$anonfun$handleTaskSetFailed$1(DAGScheduler.scala:1131)
at org.apache.spark.scheduler.DAGScheduler.$anonfun$handleTaskSetFailed$1$adapted(DAGScheduler.scala:1131)
at scala.Option.foreach(Option.scala:407)
at org.apache.spark.scheduler.DAGScheduler.handleTaskSetFailed(DAGScheduler.scala:1131)
at org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.doOnReceive(DAGScheduler.scala:2678)
at org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.onReceive(DAGScheduler.scala:2625)
at org.apache.spark.scheduler.DAGSchedulerEventProcessLoop.onReceive(DAGScheduler.scala:2613)
at org.apache.spark.util.EventLoop$$anon$1.run(EventLoop.scala:49)
at org.apache.spark.scheduler.DAGScheduler.runJob(DAGScheduler.scala:917)
at org.apache.spark.SparkContext.runJob(SparkContext.scala:2313)
at org.apache.spark.sql.execution.datasources.v2.V2TableWriteExec.writeWithV2(WriteToDataSourceV2Exec.scala:382)
... 46 more
Caused by: org.apache.spark.SparkException: Malformed records are detected in record parsing. Current parse Mode: FAILFAST. To process malformed records as null result, try setting the option 'mode' as 'PERMISSIVE'.
at org.apache.spark.sql.avro.AvroDataToCatalyst.nullSafeEval(AvroDataToCatalyst.scala:111)
at org.apache.spark.sql.catalyst.expressions.GeneratedClass$GeneratedIteratorForCodegenStage1.processNext(Unknown Source)
at org.apache.spark.sql.execution.BufferedRowIterator.hasNext(BufferedRowIterator.java:43)
at org.apache.spark.sql.execution.WholeStageCodegenExec$$anon$1.hasNext(WholeStageCodegenExec.scala:731)
at org.apache.spark.sql.execution.datasources.v2.DataWritingSparkTask$.$anonfun$run$7(WriteToDataSourceV2Exec.scala:438)
at org.apache.spark.util.Utils$.tryWithSafeFinallyAndFailureCallbacks(Utils.scala:1615)
at org.apache.spark.sql.execution.datasources.v2.DataWritingSparkTask$.run(WriteToDataSourceV2Exec.scala:477)
at org.apache.spark.sql.execution.datasources.v2.V2TableWriteExec.$anonfun$writeWithV2$2(WriteToDataSourceV2Exec.scala:385)
at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:90)
at org.apache.spark.scheduler.Task.doRunTask(Task.scala:144)
at org.apache.spark.scheduler.Task.run(Task.scala:117)
at org.apache.spark.executor.Executor$TaskRunner.$anonfun$run$9(Executor.scala:657)
at org.apache.spark.util.Utils$.tryWithSafeFinally(Utils.scala:1581)
at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:660)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
Caused by: java.lang.ArrayIndexOutOfBoundsException: -40
at org.apache.avro.io.parsing.Symbol$Alternative.getSymbol(Symbol.java:424)
at org.apache.avro.io.ResolvingDecoder.doAction(ResolvingDecoder.java:290)
at org.apache.avro.io.parsing.Parser.advance(Parser.java:88)
at org.apache.avro.io.ResolvingDecoder.readIndex(ResolvingDecoder.java:267)
at org.apache.avro.generic.GenericDatumReader.readWithoutConversion(GenericDatumReader.java:179)
at org.apache.avro.generic.GenericDatumReader.read(GenericDatumReader.java:153)
at org.apache.avro.generic.GenericDatumReader.readField(GenericDatumReader.java:232)
at org.apache.avro.generic.GenericDatumReader.readRecord(GenericDatumReader.java:222)
at org.apache.avro.generic.GenericDatumReader.readWithoutConversion(GenericDatumReader.java:175)
at org.apache.avro.generic.GenericDatumReader.read(GenericDatumReader.java:153)
at org.apache.avro.generic.GenericDatumReader.read(GenericDatumReader.java:145)
at org.apache.spark.sql.avro.AvroDataToCatalyst.nullSafeEval(AvroDataToCatalyst.scala:100)
... 16 more
Tech
NOT using the following
Ok so I just figured out what the issue was. It was in how we were generating the avro message before sending it to event hub. In our serialization method we were using the var writer = new GenericDatumWriter<GenericRecord>(schema); and IFileWriter<GenericRecord> to write to a memory stream and then just getting the byte array of that stream as seen below.
public byte[] Serialize(DataCapture data)
{
var schema = GenerateSchema(data.Schema);
var writer = new GenericDatumWriter<GenericRecord>(schema);
using(var ms = new MemoryStream())
{
using (IFileWriter<GenericRecord> fileWriter = DataFileWriter<GenericRecord>.OpenWriter(writer, ms))
{
foreach (var jsonString in data.Rows)
{
var record = new GenericRecord(schema);
var obj = JsonConvert.DeserializeObject<JObject>(jsonString);
foreach (var column in data.Schema.Columns)
{
switch (MapDataType(column.DataTypeName))
{
case AvroTypeEnum.Boolean:
record.Add(column.ColumnName, obj.GetValue(column.ColumnName).Value<bool?>());
break;
//Map all datatypes ect....removed to shorten example
default:
record.Add(column.ColumnName, obj.GetValue(column.ColumnName).Value<string>());
break;
}
}
fileWriter.Append(record);
}
}
return ms.ToArray();
}
}
When what we actually should do is use var writer = new DefaultWriter(schema); and var encoder = new BinaryEncoder(ms); to then write the records with writer.Write(record, encoder); before returning the byte array of the stream.
public byte[] Serialize(DataCapture data)
{
var schema = GenerateSchema(data.Schema);
var writer = new DefaultWriter(schema);
using (var ms = new MemoryStream())
{
var encoder = new BinaryEncoder(ms);
foreach (var jsonString in data.Rows)
{
var record = new GenericRecord(schema);
var obj = JsonConvert.DeserializeObject<JObject>(jsonString);
foreach (var column in data.Schema.Columns)
{
switch (MapDataType(column.DataTypeName))
{
case AvroTypeEnum.Boolean:
record.Add(column.ColumnName, obj.GetValue(column.ColumnName).Value<bool?>());
break;
//Map all datatypes ect....removed to shorten example
default:
record.Add(column.ColumnName, obj.GetValue(column.ColumnName).Value<string>());
break;
}
}
writer.Write(record, encoder);
}
return ms.ToArray();
}
}
So lesson learned is that not all Avro memory streams converted to byte[] are the same. The from_avro method will only de-serialize avro data the has been binary encoded with the BinaryEncoder class not data created with the IFileWriter. If there is something that I should be doing instead please let me know but this fixed my issue. Hopefully my pain will spare others the same.