Netty problem with large number of connections

[Trustin Lee (이희승)]

I agree with David that this might be the limitation of O/S or the
server might have run out of ports.

Please make sure there are not many TIME_WAIT states.

BTW, in Linux, the hard limit of the backlog value is much less than 1000 IIRC.

HTH

— Trustin Lee, http://gleamynode.net/



On Mon, Oct 26, 2009 at 9:45 PM, AD <elhoim at gmail.com> wrote:
> Hi,
>
> Could it be due to an OS limitation in the number of open connections?
> Which OS do you use?
>
> David André
>
> On Mon, Oct 19, 2009 at 12:52, Tomasz Lewandowski
> <tomasz.lewandowski at poczta.onet.pl> wrote:
>>
>> Hello,
>>
>> I have a simple HTTP Server implemented with Netty.
>>
>> I also have a stress test client (also written with Netty), which tries to
>> start 10'000 long living connections for receiving Comet-like notifications
>> (chunks of HTML in the form: "<script>....</script>", server writes about 50
>> chunks per second, so each simulated users gets a notification ones per 200
>> seconds).
>>
>> The client also generates aproximately 400 HTTP requests/connections (apart
>> from 10'000 long living connections) per second.
>>
>> Everything is fine when I run 1 instance of the stress client. But when I
>> try to run 2 instances on 2 different client machines (which would give me
>> total of 20'000 Comet-like connections), after about 18'000 connections
>> estabilished, I'm starting to get:
>>
>> java.net.ConnectException: connection timed out
>>        at
>> org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink$Boss.processConnectTimeout(NioClientSocketPipelineSink.java:361)
>>        at
>> org.jboss.netty.channel.socket.nio.NioClientSocketPipelineSink$Boss.run(NioClientSocketPipelineSink.java:273)
>>        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)
>>
>> My server works on
>> Intel(R) Core(TM)2 Duo CPU     E4500  @ 2.20GHz
>> machine with 2GB RAM which is located in the same LAN as stress clients.
>>
>> My server bootstrap code looks like this:
>>                ServerBootstrap bootstrap = new ServerBootstrap(
>>                                new NioServerSocketChannelFactory(Executors
>>                                                .newCachedThreadPool(), Executors
>>                                                .newCachedThreadPool()));
>>
>> I tried this on the server side:
>> bootstrap.setOption("backlog", 1000);
>> but it didn't help.
>>
>> Can anyone give me a hint what should I do to resolve this problem?
>>
>> Tomasz Lewandowski
>> --
>> View this message in context: http://n2.nabble.com/Netty-problem-with-large-number-of-connections-tp3848540p3848540.html
>> Sent from the Netty User Group mailing list archive at Nabble.com.
>> _______________________________________________
>> netty-users mailing list
>> netty-users at lists.jboss.org
>> https://lists.jboss.org/mailman/listinfo/netty-users
>>
>
> _______________________________________________
> netty-users mailing list
> netty-users at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/netty-users
>