[SPARK-14290][CORE][NETWORK] avoid significant memory copy in netty's transferTo

## What changes were proposed in this pull request?
When netty transfer data that is not `FileRegion`, data will be in format of `ByteBuf`, If the data is large, there will occur significant performance issue because there is memory copy underlying in `sun.nio.ch.IOUtil.write`, the CPU is 100% used, and network is very low.

In this PR, if data size is large, we will split it into small chunks to call `WritableByteChannel.write()`, so that avoid wasting of memory copy. Because the data can't be written within a single write, and it will call `transferTo` multiple times.

## How was this patch tested?
Spark unit test and manual test.
Manual test:
`sc.parallelize(Array(1,2,3),3).mapPartitions(a=>Array(new Array[Double](1024 * 1024 * 50)).iterator).reduce((a,b)=> a).length`

For more details, please refer to [SPARK-14290](https://issues.apache.org/jira/browse/SPARK-14290)

Author: Zhang, Liye <liye.zhang@intel.com>

Closes #12083 from liyezhang556520/spark-14290.
This commit is contained in:
Zhang, Liye 2016-04-06 16:11:59 -07:00 committed by Marcelo Vanzin
parent d717ae1fd7
commit c4bb02abf2

View file

@ -18,6 +18,7 @@
package org.apache.spark.network.protocol;
import java.io.IOException;
import java.nio.ByteBuffer;
import java.nio.channels.WritableByteChannel;
import javax.annotation.Nullable;
@ -43,6 +44,14 @@ class MessageWithHeader extends AbstractReferenceCounted implements FileRegion {
private final long bodyLength;
private long totalBytesTransferred;
/**
* When the write buffer size is larger than this limit, I/O will be done in chunks of this size.
* The size should not be too large as it will waste underlying memory copy. e.g. If network
* avaliable buffer is smaller than this limit, the data cannot be sent within one single write
* operation while it still will make memory copy with this size.
*/
private static final int NIO_BUFFER_LIMIT = 256 * 1024;
/**
* Construct a new MessageWithHeader.
*
@ -128,8 +137,27 @@ class MessageWithHeader extends AbstractReferenceCounted implements FileRegion {
}
private int copyByteBuf(ByteBuf buf, WritableByteChannel target) throws IOException {
int written = target.write(buf.nioBuffer());
ByteBuffer buffer = buf.nioBuffer();
int written = (buffer.remaining() <= NIO_BUFFER_LIMIT) ?
target.write(buffer) : writeNioBuffer(target, buffer);
buf.skipBytes(written);
return written;
}
private int writeNioBuffer(
WritableByteChannel writeCh,
ByteBuffer buf) throws IOException {
int originalLimit = buf.limit();
int ret = 0;
try {
int ioSize = Math.min(buf.remaining(), NIO_BUFFER_LIMIT);
buf.limit(buf.position() + ioSize);
ret = writeCh.write(buf);
} finally {
buf.limit(originalLimit);
}
return ret;
}
}