Using a client port range

["Trustin Lee (이희승)"]

Hi Neil,

ClientBootstrap.connect() does not return a boolean but returns a
ChannelFuture.  What property values does the returned ChannelFuture have?

The stack trace you pasted tells me that the exception was not triggered
by write().  If you wrote something and it triggered the exception,
NioWorker should show up in the trace instead of
NioClientSocketPipelineSink.

To make sure if this really is a Netty issue, could you please post a
very simple application that reproduces the problem?

Thanks in advance,
Trustin

Neil Avery wrote:
> Hi all,
> Ive been given a requirement to make client ports run in a particular
> range. Being the client side, it looks like java client sockets dont
> support the SO_ADDR_REUSE in that client-sockets in TIMED_WAIT cannot be
> bound to - they blow up with 'Address already in use'. 
> 
> To overcome this problem, Ive had to resort to determining availability
> by establishing a client socket connection to a server socket - not
> pretty - but using a ServerSocket() with addressResult(false) doesn't
> work - its like the flag is ignored and allows those ports to be reused.
> BTW - Im using jdk sockets for the checking process.
> 
> The problem Ive encountered is when a race-state occurs and many client
> socket connections start at the same time, it seems that the
> bootstrap.connect(serverSocket,clientSocket) - is returning 'true'
> indicating a connect success, however it turns out that the connection
> is not completed so when the channel is  used and I receive the
> exception below. I would like the ability to know the connection was
> actually successful/bound as a result of the connect future, so its
> possible to take corrective action, rather than at the write state/stage.
> 
> java.net.BindException: Address already in use: no further information
> at sun.nio.ch.SocketChannelImpl.checkConnect(Native Method)
> at sun.nio.ch.SocketChannelImpl.finishConnect(SocketChannelImpl.java:574)
> at
> org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink$Boss.connect(NioClientSocketPipelineSink.java:384)
> at
> org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink$Boss.processSelectedKeys(NioClientSocketPipelineSink.java:354)
> at
> org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink$Boss.run(NioClientSocketPipelineSink.java:276)
> at
> org.jboss.netty.util.internal.IoWorkerRunnable.run(IoWorkerRunnable.java:46)
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886)
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908)
> at java.lang.Thread.run(Thread.java:619)
> 
> Its fairly eay to reproduce.... use netty to establish a client socket
> connection on a port in a timed_wait state, and it will report
> bootstrap.connect(remote,local) == true. Then when writing to the socket
> it will fail with the above exception.
> 
> 
> Regards Neil.
> 
> -- 
> 
> www.liquidlabs.co.uk <http://www.liquidlabs.co.uk>
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> netty-users mailing list
> netty-users at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/netty-users

-- 
what we call human nature in actuality is human habit
http://gleamynode.net/