Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.java.programmer > #9278 > unrolled thread

Network Timeouts

Back to article view | Back to comp.lang.java.programmer

Contents

  Network Timeouts Tony <anthony.sterrett@gmail.com> - 2011-10-28 10:20 -0700
    Re: Network Timeouts Daniele Futtorovic <da.futt.news@laposte-dot-net.invalid> - 2011-10-29 03:15 +0200
    Re: Network Timeouts Roedy Green <see_website@mindprod.com.invalid> - 2011-10-31 03:16 -0700
      Re: Network Timeouts Andreas Leitgeb <avl@gamma.logic.tuwien.ac.at> - 2011-10-31 13:06 +0000
    Re: Network Timeouts Tony <anthony.sterrett@gmail.com> - 2011-10-31 11:03 -0700
    Re: Network Timeouts Joe Collins <joe.collins.se@gmail.com> - 2011-11-01 00:38 -0700

#9278 — Network Timeouts

Hello:
I'm writing some code between a server and a client. The client
provides the server with packets on a continuous basis. If the server
does not receive a packet for 30 seconds it throws a timeout
exception. We are testing the robustness of the server. The client is
just a producer to packets with no interleave, fixed packet length and
random context. The next packet is sent as soon as the last ack is
received. It have been observed from the logs the I get the exception
throw information (which I wrote saying the exception has been throw).
and that's about all. My approach to solving this would to be to look
at what packets were going across the interface with something like
tcpdump. But I can't get root access on this system. I can't do much
on this system in regard to taking a really close look. Any hints?
Thanks in Advance.
Cheers

[toc] | [next] | [standalone]

#9289

On 28/10/2011 19:20, Tony allegedly wrote:
> Hello:
> I'm writing some code between a server and a client. The client
> provides the server with packets on a continuous basis. If the server
> does not receive a packet for 30 seconds it throws a timeout
> exception. We are testing the robustness of the server. The client is
> just a producer to packets with no interleave, fixed packet length and
> random context. The next packet is sent as soon as the last ack is
> received. It have been observed from the logs the I get the exception
> throw information (which I wrote saying the exception has been throw).
> and that's about all. My approach to solving this would to be to look
> at what packets were going across the interface with something like
> tcpdump. But I can't get root access on this system. I can't do much
> on this system in regard to taking a really close look. Any hints?
> Thanks in Advance.
> Cheers

Not sure I understand your question. Are you asking how you could best
debug your application? Or how you could write it better so that it be
more easily debugged?

If the latter: logs, logs, logs, and JMX if you wanna be fancy.

If the former, and you do not have some fine-grained logs you can
enable, and can't sinff the traffic, then I guess you're pretty much
screwed.

-- 
DF.
Determinism trumps correctness.

[toc] | [prev] | [next] | [standalone]

#9321

On Fri, 28 Oct 2011 10:20:38 -0700 (PDT), Tony
<anthony.sterrett@gmail.com> wrote, quoted or indirectly quoted
someone who said :

>Hello:
>I'm writing some code between a server and a client. The client
>provides the server with packets on a continuous basis. If the server
>does not receive a packet for 30 seconds it throws a timeout
>exception. We are testing the robustness of the server. The client is
>just a producer to packets with no interleave, fixed packet length and
>random context. The next packet is sent as soon as the last ack is
>received. It have been observed from the logs the I get the exception
>throw information (which I wrote saying the exception has been throw).
>and that's about all. My approach to solving this would to be to look
>at what packets were going across the interface with something like
>tcpdump. But I can't get root access on this system. I can't do much
>on this system in regard to taking a really close look. Any hints?
>Thanks in Advance.
>Cheers
I use WireShark.  I don't know how it would be for you.
-- 
Roedy Green Canadian Mind Products
http://mindprod.com
It's difficult to be rigorous about whether a machine really knows,
thinks, etc., because we’re hard put to define these things. 
We understand human mental processes only slightly better than
a fish understands swimming. 
~ John McCarthy (born: 1927-09-04 died: 2011-10-23 at age: 84).
Inventor of the term AI (Artificial Intelligence), 
the short-circuit OR operator (|| in Java), 
and LISP (LIst Processing Language) that makes EMACS 
(Extensible MACro System) so addictive.

[toc] | [prev] | [next] | [standalone]

#9327

Roedy Green <see_website@mindprod.com.invalid> wrote:
>>I'm writing some code between a server and a client. The client
>>provides the server with packets on a continuous basis. If the server
>>does not receive a packet for 30 seconds it throws a timeout
>>exception. We are testing the robustness of the server. The client is
>>just a producer to packets with no interleave, fixed packet length and
>>random context. The next packet is sent as soon as the last ack is
>>received. It have been observed from the logs the I get the exception
>>throw information (which I wrote saying the exception has been throw).
>>and that's about all. My approach to solving this would to be to look
>>at what packets were going across the interface with something like
>>tcpdump. But I can't get root access on this system. I can't do much
>>on this system in regard to taking a really close look. Any hints?
>>Thanks in Advance.
>>Cheers
> I use WireShark.  I don't know how it would be for you.

WireShark needs root-privileges, as well, so probably not.

to the OP:
  How about changing the exeception handler to write some 
  more useful information into the logs?

[toc] | [prev] | [next] | [standalone]

#9329

On Oct 28, 1:20 pm, Tony <anthony.sterr...@gmail.com> wrote:
> Hello:
> I'm writing some code between a server and a client. The client
> provides the server with packets on a continuous basis. If the server
> does not receive a packet for 30 seconds it throws a timeout
> exception. We are testing the robustness of the server. The client is
> just a producer to packets with no interleave, fixed packet length and
> random context. The next packet is sent as soon as the last ack is
> received. It have been observed from the logs the I get the exception
> throw information (which I wrote saying the exception has been throw).
> and that's about all. My approach to solving this would to be to look
> at what packets were going across the interface with something like
> tcpdump. But I can't get root access on this system. I can't do much
> on this system in regard to taking a really close look. Any hints?
> Thanks in Advance.
> Cheers

Thanks all

[toc] | [prev] | [next] | [standalone]

#9331

If you can't get root access or otherwise use wireshark, have you
considered setting up a proxy?  All it needs to do is write the
packets to disk and then forward them along.

On Oct 28, 5:20 pm, Tony <anthony.sterr...@gmail.com> wrote:
> Hello:
> I'm writing some code between a server and a client. The client
> provides the server with packets on a continuous basis. If the server
> does not receive a packet for 30 seconds it throws a timeout
> exception. We are testing the robustness of the server. The client is
> just a producer to packets with no interleave, fixed packet length and
> random context. The next packet is sent as soon as the last ack is
> received. It have been observed from the logs the I get the exception
> throw information (which I wrote saying the exception has been throw).
> and that's about all. My approach to solving this would to be to look
> at what packets were going across the interface with something like
> tcpdump. But I can't get root access on this system. I can't do much
> on this system in regard to taking a really close look. Any hints?
> Thanks in Advance.
> Cheers

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.java.programmer

csiph-web