Thanks tatu,
Did you try with client on linux as well? Did you get the same problem?
What are the workarounds you talk about? I couldn't find them?
Thanks in advance
Pascal
-----Original Message-----
From: Tatu Saloranta [mailto:tsaloranta_at_gmail.com]
Sent: Wednesday, June 10, 2009 2:00 PM
To: users_at_jersey.dev.java.net
Subject: Re: [Jersey] Weird exception: java.lang.BindException: Address
already in use.
I can confirm this as well -- I once tried to do load testing for
Jetty-web-service on Linux, client on windows (to prove you can do up
to 10k tps on regular work station level machine, servlets+xml).
And sure enough, after maybe 10 seconds you start getting exceptions
due to problems with releasing sockets (10 seconds since some http
connections can be reused with 1.1 connection reuse).
There are some workarounds for this, but I assume partly it's a
conscious decision by MS, not to make 'consumer' windows too tempting
platform for development. :-)
(no comment on what else is probably involved :-D ).
-+ Tatu +-
On Wed, Jun 10, 2009 at 6:45 AM, Gehl, Pascal<Pascal.Gehl_at_nuance.com>
wrote:
> Thanks a lot Tim,
>
> After I read your answer I did some research and it's true Windows
does
> have problem "releasing" ports under load.
>
> Pascal
>
> -----Original Message-----
> From: Tim Edwards [mailto:Edwards.T_at_cambridgeassessment.org.uk]
> Sent: Wednesday, June 10, 2009 4:11 AM
> To: users_at_jersey.dev.java.net
> Subject: RE: [Jersey] Weird exception: java.lang.BindException:
Address
> already in use.
>
>
> Hi,
>
> Not sure if it is the cause here, but I have experienced this
exception
> on the client side of one of our projects (although it does not use
> Jersey). The basic problem is, if running on Windows, it will only
allow
> a certain number of ports to be open simultaneously and if you exceed
> that limit then you will receive that exception.
>
> We experienced it when running a suite of integration tests with 3000
> odd cases, although we were not using that may ports at once, it takes
> time for windows to recycle the port once you have finished with it,
so
> a lot of ports get left in a unusable state and eventually run out. It
> was a random exception that didn't always happen or happen in the same
> place but you could see all the ports being used by running a netstat.
>
> We managed to alleviate the problem slightly through better use of
> HttpClient but it has not solved it completely. You can set the
maximum
> port limit in Windows via the registry but that's not a particularly
> nice "fix".
>
> Cheers,
> Tim
>
>
>
> -----Original Message-----
> From: Paul.Sandoz_at_Sun.COM [mailto:Paul.Sandoz_at_Sun.COM]
> Sent: 10 June 2009 05:29
> To: users_at_jersey.dev.java.net
> Subject: Re: [Jersey] Weird exception: java.lang.BindException:
Address
> already in use.
>
>
> On Jun 9, 2009, at 7:28 PM, Gehl, Pascal wrote:
>
>> Thanks for the answer.
>>
>> But this exception occurs in my already running instance on the
client
>
>> side when the client API tries to send a HTTP post, not at startup.
>>
>
> I misunderstood your original email, when you were describing the
> similarity.
>
> I see the exception occurs when the client is attempting to open the
> connection, even before any bytes are written.
>
> Are you saying this only occurs with a POST request? if so what about
a
> GET request?
>
> How many requests do you make for this to occur?
>
>
> It is indeed an odd exception on the client side. I have never
observed
> this before (we have some tests that start embedded servers and use
the
> client API). It is not clear to me if the error message is being sent
> from the server to the client, i would presume it would be since this
is
> really associated with the server side, implying some issue with the
> server side configuration as Tatu suggests.
>
> Paul.
>
>> Pascal
>>
>> -----Original Message-----
>> From: Tatu Saloranta [mailto:tsaloranta_at_gmail.com]
>> Sent: Tuesday, June 09, 2009 12:10 PM
>> To: users_at_jersey.dev.java.net
>> Subject: Re: [Jersey] Weird exception: java.lang.BindException:
>> Address
>> already in use.
>>
>> On Tue, Jun 9, 2009 at 8:32 AM, Gehl, Pascal<Pascal.Gehl_at_nuance.com>
>> wrote:
>>> I don't have 2 servers listening on the same port on the same
>>> machine.
>>
>> I wouldn't be so sure. Exception basically says that yes, there is
>> already a process that listens to the port.
>>
>> The most common case is that you already have a running instance --
>> possibly one that has not cleanly shut down when it should have --
and
>
>> try to start a new one. It could be due to startup script not
ensuring
>
>> previous instance is not properly shut down, or not waiting for that
>> to occur.
>>
>> Btw, this is a general web server startup problem, and quite possibly
>> not specific to Jersey. You can try googling for it, or check out
your
>
>> servlet container's FAQ, since it is a very common issue.
>>
>> Hope this helps,
>>
>> -+ Tatu +-
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: users-unsubscribe_at_jersey.dev.java.net
>> For additional commands, e-mail: users-help_at_jersey.dev.java.net
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: users-unsubscribe_at_jersey.dev.java.net
>> For additional commands, e-mail: users-help_at_jersey.dev.java.net
>>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe_at_jersey.dev.java.net
> For additional commands, e-mail: users-help_at_jersey.dev.java.net
>
>
>
> Click
> https://www.mailcontrol.com/sr/wQw0zmjPoHdJTZGyOCrrhg==
> h3sjGTelFxQnR1Q7l6e!9bxE+P8w3h9RYiELIeSJhirLg== to report this email
as
> spam.
>
>
> If you are not the intended recipient, employee or agent responsible
for
> delivering the message to the intended recipient, you are hereby
> notified that any dissemination or copying of this communication and
its
> attachments is strictly prohibited.
> If you have received this communication and its attachments in error,
> please return the original message and attachments to the sender using
> the reply facility on e-mail.
> Internet communications are not secure and therefore Cambridge
> Assessment (the brand name for the University of Cambridge Local
> Examinations Syndicate, the constituent elements of which are CIE,
ESOL
> and OCR [Oxford Cambridge and RSA Examinations is a Company Limited by
> Guarantee Registered in England. Registered office: 1 Hills Road,
> Cambridge CB1 2EU. Company number: 3484466]) does not accept legal
> responsibility for the contents of this message.
> Any views or opinions presented are solely those of the author and do
> not necessarily represent those of Cambridge Assessment unless
otherwise
> specifically stated.
> The information contained in this email may be subject to public
> disclosure under the Freedom of Information Act 2000. Unless the
> information is legally exempt from disclosure, the confidentiality of
> this email and your reply cannot be guaranteed.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe_at_jersey.dev.java.net
> For additional commands, e-mail: users-help_at_jersey.dev.java.net
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe_at_jersey.dev.java.net
> For additional commands, e-mail: users-help_at_jersey.dev.java.net
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe_at_jersey.dev.java.net
For additional commands, e-mail: users-help_at_jersey.dev.java.net