On Mon, 2001-10-29 at 20:31, Chemolli Francesco (USI) wrote:
> > On Mon, 2001-10-29 at 09:54, Tony Melia wrote:
> > > there
> > > are lots of anonymous TCP_DENIED which seem to be
> > generated/required by
> > > fakeauth, followed by a successful log with the username recorded.
> >
> > The extra TCP_DENIED's are microsoft's wonderful legacy. We've
> > considered filterting them out, but until we can be sure there is no
> > lost information of a useful nature, tjey will remain.
>
> As far as I am concerned they _will_ remain, at least to record
> the incredible amount of bandwidth wasted.
Once we get chunked encoding the overhead should drop significantly.
Still, I'm not in favour of completely eliminating the middle deny (the
first deny must stay, we cannot identify it as being in-progress), but
rather adding a configure option to allow folk who only look at the
request stats, not the byte stats to get semi-sensible info.
Rob
Received on Mon Oct 29 2001 - 02:28:58 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:03:11 MST