Henrik Nordstrom wrote:
> tis 2006-05-23 klockan 13:22 -0400 skrev Nick Lewycky:
>
>
>>It's a temporary. pe->getReply()->head will get casted into the
>>HttpHeader, used to call putStr(...), then its life is over, at which
>>point the C++ implementation is f
ent, some wait until the end of
the function).
Even if you'd said "HttpHeader HH = ...", it would still create a
temporary, copy-construct the new HttpHeader, and then destroy the
temporary.
Nick Lewycky
start using it. That'd be a start. Beyond
that, finish the transitions I guess. Try to make Squid a more
straight-forward pipeline.
Sorry if I sound too negative. I'm trying to be helpful with this email,
but I find it very difficult to point at one thing and say "there's your
problem". As for me, I want squid-prefetching off my hands, whether by
landing it into HEAD or just walking away from it.
Nick Lewycky
Leandro Scott R.Z. Jacques wrote:
> I'm needing some tool to analyse squid's access.log to
> check cache hits and cache misses and how many bytes
> had to be downloaded from the origin server due to the
> misses and how many bytes was returned from the cache
> by cache hits. Is there any tool that
would remove RefCount<> first, then hopefully CBDATA in the future.
The nice thing about using Boost is that if the library is added to C++
in the (admittedly distant) future, our API would not need to change,
and our copy of the Boost implementation could be kept around for older
compilers.
Nick Lewycky
Jon Kay wrote:
> Nick Lewycky wrote:
>
>>Finally, does anyone have suggestions for how to test for performance
>>improvement due to prefetching?
>
> A good way to test how your algorithms are working is to get a nice, long
> actual Squid workload -eg, URLs fetched, an
Alex Rousskov wrote:
> On Thu, 2005/05/12 (MDT), <[EMAIL PROTECTED]> wrote:
>
>> does anyone have suggestions for how to test for performance
>> improvement due to prefetching?
>
> You can use Content Simulation Module (CSM) with Web Polygraph. It lets
> you benchmark using real content, includin
d it
(after analyzing it for more prefetchables if it's HTML).
Finally, does anyone have suggestions for how to test for performance
improvement due to prefetching?
Thanks,
Nick Lewycky
Henrik Nordstrom wrote:
On Fri, 11 Mar 2005, Nick Lewycky wrote:
is a branch called "collapsed forwarding"[4], but it currently only
applies to accelerator setups. It could perhaps be adapted.
It appliest equal to all setups. It was designed for accelerator setups
as there is where th
x27;t dealt with this yet and we'll both
need a solution. There is a branch called "collapsed forwarding"[4], but
it currently only applies to accelerator setups. It could perhaps be
adapted.
Good luck.
Nick Lewycky
[1] - http://squidwiki.kinkie.it/squidwiki/ClientStreams
[2] - htt
theory, it should never show worse numbers.
If this is your project, I'd very much appreciate an extra pair of
hands. There's a lot of work left to do.
Thanks,
Nick Lewycky
and I need a way to simulate additional network
latency. This is the most promising solution I know of at the moment,
but if anyone has a different solution (such as Linux netfilter tricks?)
I can use, that would work just as well.
Thanks,
Nick Lewycky
tch it up, but the constructor probably ought
to belong a cc file instead of the header.
Developers making these sorts of changes might want to test them with a
tool like Valgrind from valgrind.kde.org (Linux-specific, sorry) which
will detect dependent reads to uninitialized memory.
The following fragment is needed to compile Squid with gcc-3.4:
Index: include/Range.h
===
RCS file: /cvsroot/squid/squid3/include/Range.h,v
retrieving revision 1.5
diff -u -w -u -r1.5 Range.h
--- include/Range.h 23 Sep 2003 02:12:
lient
stream callbacks? Are there any plans to port http.cc over to the client
stream API? If not, is there some fundamental reason not to that I've
missed?
Nick Lewycky
15 matches
Mail list logo