Re: (FWD) Results of stress testing the current version of the Onion Router

[Bruce Montrose <montrose@itd.nrl.navy.mil>]

That fixed the problem Roger. Performance under stress is still pretty bad, 
but at least it doesn't crash now when I beat it up :)

I put the latest results of the stress testing in a tarball called 
pvm.tar.gz at http://freehaven.net/~montrose/

At 01:47 PM 9/3/2002 -0400, you wrote:
>On Tue, Sep 03, 2002 at 12:56:19PM -0400, Bruce Montrose wrote:
> > I had to kill off both the httpap and the OR on the first machine and then
> > restart them both before it would start working again.
>
>Gosh.
>Confirmed that it's a bug in the OR. It happens when we get to maxconn
>connections; something goes haywire and it stops listening at the sockets.
>I'll track it down.
>
> > However, I suspect that the output buffers may not have been flushed on 
> the
> > processes that hung. Does anyone know how to make a process not buffer its
> > output so that it gets written immediately?
>
>What kind of process are we talking about here? The onion routers
>explicitly buffer onto a string, and only attempt to flush the buffer
>periodically.
>
>--Roger