Re:Re: Sorry for interrupting again :-(

Thanks, Joe pal!

I will follow what you said. I am a newbie, if I ask silly questions, I will
say sorry. :-)

Best regards,
George, Ma

----- Original Message -----
From: Joe Cooper
To: maer727@sohu.com ;squid-users@squid-cache.org
Subject: Re: Sorry for interrupting again :-(
Sent: Thu Apr 11 10:58:08 CST 2002

> You /can't/ do that!
>
> But the best tool for you to use is Polygraph. Go get it. Read the
> docs. Use it.
>
> I promise you polygraph is what you want. There is no use asking more
> questions about getting URLs from cacheoffs (this is the fourth time
> you've asked and the third time I've explained why you can't go about
> your project that way).
>
> But you /can't/ get a list of URLs from the cacheoffs and use them for
> your tests. That isn't how polygraph works. So again, I implore you:
> Go get polygraph. Read the docs. Use it.
>
> maer727@sohu.com wrote:
> > Thanks, Joe pal!
> >
> > I think PolyGraph is a nice tool for me. But I am look for a tool that can
> > use cacheoffs as a source of URLs to run through my cache. Which is the best
> > tool for me to use?
> >
> > Best regards,
> > George, Ma
> >
> >
> > ----- Original Message -----
> > From: Joe Cooper
> > To: maer727@sohu.com
> > Cc: squid-dev@squid-cache.org
> > Subject: Re: Sorry for interrupting again :-(
> > Sent: Wed Apr 10 22:33:09 CST 2002
> >
> >
> >>I don't think I can make it much clearer. You can't use the Polygraph
> >>logs from cacheoffs as a source of URLs to run through your cache. If
> >>you understood the other sentences surrounding that one, then you got my
> >>meaning already (I repeated the concept several times, since I could
> >>tell you were having trouble grasping what I was talking about).
> >>
> >>maer727@sohu.com wrote:
> >>
> >>>Hi, Joe pal!
> >>>
> >>>Sorry for interrupting again. I re-read your reply. And find a question.
> >>>You mentioned, (the last paragrath)
> >>>
> >>>///////////////////////////////////////////////////////////////////
> >>>not for simulating a Polygraph benchmark run using a different utility
> >>>///////////////////////////////////////////////////////////////////
> >>>
> >>>What does that mean?
> >>>
> >>>Best regards,
> >>>George, Ma
> >>>
> >>>----- Original Message -----
> >>>From: Joe Cooper
> >>>To: maer727@sohu.com
> >>>Cc: squid-dev@squid-cache.org
> >>>Subject: Re: what means Bake-off?
> >>>Sent: Wed Apr 10 18:20:14 CST 2002
> >>>
> >>>
> >>>
> >>>>Because they are not traces of 'internet' activity. They are the result
> >>>>of benchmark run that uses entirely artificial URLs. Every run is
> >>>>different, and every run is generated on the fly by the polygraph server
> >>>>and client. They aren't useful for your purposes because you requested
> >>>>a list of URLs that you could run through your Squid to test hit ratios.
> >>>> The URLs generated during a Polygraph run do not resolve to a real
> >>>>internet host--none of them exist on the internet.
> >>>>
> >>>>Get Polygraph (it's free). Read the docs for it. Try it. You'll then
> >>>>understand what I'm talking about.
> >>>>
> >>>>What I'm getting at is that polygraph will be extremely useful for you.
> >>>> There is no better cache benchmarking tool. /But/ the logs it
> >>>>generates are only useful for creating benchmark result reports--not for
> >>>>simulating a Polygraph benchmark run using a different utility. If you
> >>>>want to reproduce a cache-off style benchmark, you've got to use
> >>>>Polygraph. Logs from the cache-offs are useless unless you want to
> >>>>analyze the results of servers that took part in the cache-offs--they
> >>>>are not useful for reproducing the same conditions.
> >>>>
> >>>>maer727@sohu.com wrote:
> >>>>
> >>>>
> >>>>>Thanks, Joe pal!
> >>>>>
> >>>>>In your reply, you mentioned,
> >>>>>
> >>>>>////////////////////////////////////////////////
> >>>>>only logs produced are the logs that are generated during the
> >>>>>benchmark runs,
> >>>>>////////////////////////////////////////////////
> >>>>>
> >>>>>I do not understand the meaning above, I am a newbie. :-) Can you give
> >>>>>me a simple explanation? Why these logs are not useful for my purpose?
> >>>>>
> >>>>>Best regards,
> >>>>>George, Ma
> >>>>>
......
Received on Wed Apr 10 2002 - 21:19:31 MDT