Clarify reason why SqlSerializer2 supports this serializer

jeanlyn · May 5, 2015 · 0a7ebd7 · 0a7ebd7
1 parent 123b992
commit 0a7ebd7
Show file tree

Hide file tree

Showing 2 changed files with 6 additions and 2 deletions.
diff --git a/core/src/main/scala/org/apache/spark/serializer/Serializer.scala b/core/src/main/scala/org/apache/spark/serializer/Serializer.scala
@@ -86,7 +86,8 @@ abstract class Serializer {
    * serIn.open([obj2bytes] concatenate [obj1bytes]) should return (obj2, obj1)
    * }}}
    *
-   * In general, this property should hold for serializers that are stateless.
+   * In general, this property should hold for serializers that are stateless and that do not
+   * write special metadata at the beginning or end of the serialization stream.
    *
    * This API is private to Spark; this method should not be overridden in third-party subclasses
    * or called in user code and is subject to removal in future Spark releases.

diff --git a/sql/core/src/main/scala/org/apache/spark/sql/execution/SparkSqlSerializer2.scala b/sql/core/src/main/scala/org/apache/spark/sql/execution/SparkSqlSerializer2.scala
@@ -155,7 +155,10 @@ private[sql] class SparkSqlSerializer2(keySchema: Array[DataType], valueSchema:
 
   def newInstance(): SerializerInstance = new ShuffleSerializerInstance(keySchema, valueSchema)
 
-  override def supportsRelocationOfSerializedObjects: Boolean = true
+  override def supportsRelocationOfSerializedObjects: Boolean = {
+    // SparkSqlSerializer2 is stateless and writes no stream headers
+    true
+  }
 }
 
 private[sql] object SparkSqlSerializer2 {