SerializerManager¶

SerializerManager is used to <> for shuffle blocks (either the default <> or <> based on the key and value of a record).

== [[creating-instance]] Creating Instance

SerializerManager takes the following to be created:

• [[defaultSerializer]] serializer:Serializer.md[Serializer]
• [[conf]] ROOT:SparkConf.md[SparkConf]
• [[encryptionKey]] Optional encryption key ([Array[Byte]])

SerializerManager is created when SparkEnv utility is used to core:SparkEnv.md#create[create a SparkEnv for the driver and executors].

== [[SparkEnv]] Accessing SerializerManager Using SparkEnv

SerializerManager is available using core:SparkEnv.md#serializerManager[SparkEnv] on the driver and executors.

[source, scala]¶

import org.apache.spark.SparkEnv SparkEnv.get.serializerManager

== [[kryoSerializer]] KryoSerializer

SerializerManager creates a KryoSerializer when created.

KryoSerializer is used as a <> when the type of a given key and value is <>.

== [[wrapForCompression]] Wrapping Input or Output Stream of Block for Compression

[source, scala]¶

wrapForCompression( blockId: BlockId, s: OutputStream): OutputStream wrapForCompression( blockId: BlockId, s: InputStream): InputStream

wrapForCompression...FIXME

wrapForCompression is used when:

• SerializerManager is requested to <>, <>, <> and <>

• SerializedValuesHolder (of storage:MemoryStore.md[MemoryStore]) is requested for a SerializationStream

== [[wrapStream]] Wrapping Input or Output Stream for Block

[source, scala]¶

wrapStream( blockId: BlockId, s: InputStream): InputStream wrapStream( blockId: BlockId, s: OutputStream): OutputStream

wrapStream...FIXME

wrapStream is used when:

• BlockStoreShuffleReader is requested to shuffle:BlockStoreShuffleReader.md#read[read combined records for a reduce task]

• DiskMapIterator (of shuffle:ExternalAppendOnlyMap.md[ExternalAppendOnlyMap]) is requested for nextBatchStream

• SpillReader (of shuffle:ExternalSorter.md[ExternalSorter]) is requested for nextBatchStream

• memory:UnsafeSorterSpillReader.md[UnsafeSorterSpillReader] is created

• DiskBlockObjectWriter is requested to storage:DiskBlockObjectWriter.md#open[open]

== [[dataSerializeStream]] dataSerializeStream Method

[source, scala]¶

dataSerializeStreamT: ClassTag: Unit

dataSerializeStream...FIXME

dataSerializeStream is used when BlockManager is requested to storage:BlockManager.md#doPutIterator[doPutIterator] and storage:BlockManager.md#dropFromMemory[dropFromMemory].

== [[dataSerializeWithExplicitClassTag]] dataSerializeWithExplicitClassTag Method

[source, scala]¶

dataSerializeWithExplicitClassTag( blockId: BlockId, values: Iterator[], classTag: ClassTag[]): ChunkedByteBuffer

dataSerializeWithExplicitClassTag...FIXME

dataSerializeWithExplicitClassTag is used when BlockManager is requested to storage:BlockManager.md#doGetLocalBytes[doGetLocalBytes].

== [[dataDeserializeStream]] dataDeserializeStream Method

[source, scala]¶

dataDeserializeStreamT (classTag: ClassTag[T]): Iterator[T]

dataDeserializeStream...FIXME

dataDeserializeStream is used when:

• BlockManager is requested to storage:BlockManager.md#getLocalValues[getLocalValues], storage:BlockManager.md#getRemoteValues[getRemoteValues] and storage:BlockManager.md#doPutBytes[doPutBytes]

• MemoryStore is requested to storage:MemoryStore.md#putIteratorAsBytes[putIteratorAsBytes] (when PartiallySerializedBlock is requested for a PartiallyUnrolledIterator)

== [[getSerializer]] Selecting Serializer

[source, scala]¶

getSerializer( ct: ClassTag[], autoPick: Boolean): Serializer getSerializer( keyClassTag: ClassTag[], valueClassTag: ClassTag[_]): Serializer

getSerializer returns the <> when the given arguments are <>. Otherwise, getSerializer returns the <>.

getSerializer is used when:

• ShuffledRDD is requested for rdd:ShuffledRDD.md#getDependencies[dependencies].

• SerializerManager is requested to <>, <> and <>

• SerializedValuesHolder (of storage:MemoryStore.md[MemoryStore]) is requested for a SerializationStream

== [[canUseKryo]] Checking Whether Kryo Serializer Could Be Used

[source, scala]¶

canUseKryo( ct: ClassTag[_]): Boolean

canUseKryo is true when the given ClassTag is a primitive, an array of primitives or a String. Otherwise, canUseKryo is false.

canUseKryo is used when SerializerManager is requested for a <>.

== [[shouldCompress]] shouldCompress Method

[source, scala]¶

shouldCompress( blockId: BlockId): Boolean

shouldCompress...FIXME

shouldCompress is used when SerializerManager is requested to <>.