[SPARK-40819] Parquet INT64 (TIMESTAMP(NANOS,true)) now throwing Illegal Parquet type instead of automatically converting to LongType - ASF JIRA

XML

Word

Printable

JSON

Details

Type: Bug
Status: Resolved
Priority: Critical
Resolution: Fixed
Affects Version/s: 3.2.0, 3.2.1, 3.3.0, 3.2.2, 3.3.1, 3.2.3, 3.3.2, 3.4.0
Fix Version/s: 3.2.4, 3.3.2, 3.4.0
Component/s: SQL
Labels:
- regression

Language:
- scala

Description

Since 3.2 parquet files containing attributes with type "INT64 (TIMESTAMP(NANOS, true))" are no longer readable and attempting to read throws:

Caused by: org.apache.spark.sql.AnalysisException: Illegal Parquet type: INT64 (TIMESTAMP(NANOS,true))
  at org.apache.spark.sql.errors.QueryCompilationErrors$.illegalParquetTypeError(QueryCompilationErrors.scala:1284)
  at org.apache.spark.sql.execution.datasources.parquet.ParquetToSparkSchemaConverter.illegalType$1(ParquetSchemaConverter.scala:105)
  at org.apache.spark.sql.execution.datasources.parquet.ParquetToSparkSchemaConverter.convertPrimitiveField(ParquetSchemaConverter.scala:174)
  at org.apache.spark.sql.execution.datasources.parquet.ParquetToSparkSchemaConverter.convertField(ParquetSchemaConverter.scala:90)
  at org.apache.spark.sql.execution.datasources.parquet.ParquetToSparkSchemaConverter.$anonfun$convert$1(ParquetSchemaConverter.scala:72)
  at scala.collection.TraversableLike.$anonfun$map$1(TraversableLike.scala:238)
  at scala.collection.Iterator.foreach(Iterator.scala:941)
  at scala.collection.Iterator.foreach$(Iterator.scala:941)
  at scala.collection.AbstractIterator.foreach(Iterator.scala:1429)
  at scala.collection.IterableLike.foreach(IterableLike.scala:74)
  at scala.collection.IterableLike.foreach$(IterableLike.scala:73)
  at scala.collection.AbstractIterable.foreach(Iterable.scala:56)
  at scala.collection.TraversableLike.map(TraversableLike.scala:238)
  at scala.collection.TraversableLike.map$(TraversableLike.scala:231)
  at scala.collection.AbstractTraversable.map(Traversable.scala:108)
  at org.apache.spark.sql.execution.datasources.parquet.ParquetToSparkSchemaConverter.convert(ParquetSchemaConverter.scala:66)
  at org.apache.spark.sql.execution.datasources.parquet.ParquetToSparkSchemaConverter.convert(ParquetSchemaConverter.scala:63)
  at org.apache.spark.sql.execution.datasources.parquet.ParquetFileFormat$.$anonfun$readSchemaFromFooter$2(ParquetFileFormat.scala:548)
  at scala.Option.getOrElse(Option.scala:189)
  at org.apache.spark.sql.execution.datasources.parquet.ParquetFileFormat$.readSchemaFromFooter(ParquetFileFormat.scala:548)
  at org.apache.spark.sql.execution.datasources.parquet.ParquetFileFormat$.$anonfun$mergeSchemasInParallel$2(ParquetFileFormat.scala:528)
  at scala.collection.immutable.Stream.map(Stream.scala:418)
  at org.apache.spark.sql.execution.datasources.parquet.ParquetFileFormat$.$anonfun$mergeSchemasInParallel$1(ParquetFileFormat.scala:528)
  at org.apache.spark.sql.execution.datasources.parquet.ParquetFileFormat$.$anonfun$mergeSchemasInParallel$1$adapted(ParquetFileFormat.scala:521)
  at org.apache.spark.sql.execution.datasources.SchemaMergeUtils$.$anonfun$mergeSchemasInParallel$2(SchemaMergeUtils.scala:76)

Prior to 3.2 successfully reads the parquet automatically converting to a LongType.

I believe work part of https://issues.apache.org/jira/browse/SPARK-34661 introduced the change in behaviour, more specifically here: https://github.com/apache/spark/pull/31776/files#diff-3730a913c4b95edf09fb78f8739c538bae53f7269555b6226efe7ccee1901b39R154 which throws the QueryCompilationErrors.illegalParquetTypeError