org.apache.hadoop.net

Class SocketInputWrapper

java.lang.Object

java.io.InputStream

java.io.FilterInputStream

org.apache.hadoop.net.SocketInputWrapper

All Implemented Interfaces:

Closeable, AutoCloseable

@InterfaceAudience.LimitedPrivate(value="HDFS")
@InterfaceStability.Unstable
public class SocketInputWrapper
extends FilterInputStream

A wrapper stream around a socket which allows setting of its timeout. If the socket has a channel, this uses non-blocking IO via the package-private SocketInputStream implementation. Otherwise, timeouts are managed by setting the underlying socket timeout itself.

Field Summary

Fields inherited from class java.io.FilterInputStream

in

Method Summary

Methods

Modifier and Type	Method and Description
ReadableByteChannel	getReadableByteChannel()
void	setTimeout(long timeoutMs) Set the timeout for reads from this stream.

Methods inherited from class java.io.FilterInputStream

available, close, mark, markSupported, read, read, read, reset, skip

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Detail

setTimeout

public void setTimeout(long timeoutMs)
                throws SocketException

Set the timeout for reads from this stream. Note: the behavior here can differ subtly depending on whether the underlying socket has an associated Channel. In particular, if there is no channel, then this call will affect the socket timeout for all readers of this socket. If there is a channel, then this call will affect the timeout only for this stream. As such, it is recommended to only create one SocketInputWrapper instance per socket.

Parameters:

timeoutMs - the new timeout, 0 for no timeout

Throws:

SocketException - if the timeout cannot be set

getReadableByteChannel

public ReadableByteChannel getReadableByteChannel()

Returns:

an underlying ReadableByteChannel implementation.

Throws:

IllegalStateException - if this socket does not have a channel