Thus spake Sami Farin (safari-dns@safari.iki.fi):
> Well BIND seems to do better, according to this.
> http://www.lurhq.com/cachepoisoning.html
Someone else already proved you wrong on this.
> > You have provided NO evidence whatsoever that dnscache was actually
> > bottlenecked by the divisions or by surf (which is a ludicrous thought
> I showed dns_random took most of the time in one email.
So, now that you proved beyond reasonable doubt that you don't have a
clue what you are talking about, how about you go trolling somewhere
else?
> > given that it's doing socket I/O, i.e. syscalls, all the time).
> > You didn't even have evidence that poll is a bottleneck. On a typical
> Yes I have.
No you didn't.
> And epoll was not the only improvement.
It is not even clear whether it is an improvement at all. On my
production servers dnscache produces negligible load in the first place.
> Go troll somewhere else.
Bwahahaha
Your mom must be so proud of you. How old are you again? 12?
> I can make it as bad or good as I want to,
> and I don't need your opinion while doing so.
Please crawl back under your rock and do it there.
> > crappy chat protocol.
> Actually, no need to, because there's SILC, invented by a Finn.
> http://silcnet.org/
Yeah and look how crappy it is. Astonishingly so.
I'm impressed. You actually understood one thing out of five I said.
Still no cigar. Please go somewhere else. Maybe you could go to the
SILC project, these people really need the help of someone with your
level of technical expertise.
Felix
|