in reply to Re: Network Reliability Testing, problems with buffered sockets?
in thread Network Reliability Testing, problems with buffered sockets?

Oddly enough I've come across something similar before, except it was with a spam filter labelling something as spam, then the server rejecting the unknown header and sending it back to the spam filter, which labelled it as spam and returned it again.

I did think about using packet sniffing/logging to analyse the flow, but unfortunately it wouldn't give me enough information, and I'd need to work out how to track the connection.

Thanks for the idea :)

  • Comment on Re^2: Network Reliability Testing, problems with buffered sockets?

Replies are listed 'Best First'.
Re^3: Network Reliability Testing, problems with buffered sockets?
by 5mi11er (Deacon) on Mar 21, 2007 at 16:25 UTC
    I think maybe you're in tunnel vision mode with the problem. It sounds like you're assuming there's something wrong with a single application's packet flows. I'm betting there's a better chance that there is other traffic that is using enough bandwidth to keep the one application from responding well. A sniffer would allow you to see all the traffic, not just the one application you're having issues with.

    -Scott