You are viewing a plain text version of this content. The canonical link for it is here.
Posted to modperl@perl.apache.org by Adi <ad...@certsite.com> on 2000/04/18 23:41:57 UTC
[OT] LWP: Regression testing a mod_perl server
I am trying to put together a comprehensive regression suite for CGI based
mod_perl web applications using LWP. It involves sequences of CGI calls
with expected results after each consecutive request. But I noticed
something about LWP: it caches requests, almost like it expects *only*
static content. So if you loop over a given sequence of HTTP requests, only
the first one actually gets sent to the server.
eg:
foreach (1..5) {
$ua->request(GET ('http://localhost/uri1'));
$ua->request(GET ('http://localhost/uri2'));
$ua->request(GET ('http://localhost/uri3'));
}
# this code will only send 3 requests to the webserver, NOT 15
Has anyone done similar work making a regression test using LWP ? Is there
an easy way to turn off this caching behavior?
-Adi
Re: [OT] LWP: Regression testing a mod_perl server
Posted by Stas Bekman <sb...@stason.org>.
On Tue, 18 Apr 2000, Adi wrote:
> Cool, thanks for the quick response, Stas.
>
> I don't seem to have LWP::Parallel. Looks like I need to upgrade to the
> newest version of LWP. But yes, I did notice that ab gives much more
> accurate results. I was using Time::HiRes to simply measure the time it
> takes from request to response, and it gave widely varying results. Which
> was why I wanted to do a loop so I could divide the total time taken by the
> number of iterations.
Well, it's almost useless to benchmark by running a single query. I don't
know how did you want to do that without LWP::Parallel.
> I'm glad to hear you're already working on it. I'll definitely help out
> with your regression suite, because I haven't gotten too far on mine (I
> just started working on it today).
I've started 2 days ago when I've got tired from running the benchmarks
for the Guide manually and parse the results... Hold on and you might get
my version very soon now :)
>
> Thanks for the LWP list ref.. I'll post the question there.
>
> -Adi
>
> Stas Bekman wrote:
> >
> > First I work on a similar suite as well. I didn't want to announce it yet,
> > before I get some things working. But since you've mentioned it here, here
> > we go... Hope to get the first version out of the door in a few days.
> >
> > Regarding LWP, I've used a similar to ab script as described at
> > http://perl.apache.org/guide/performance.html and I've found the results
> > are very different if you the same test with ab and LWP::Parallel, so I'd
> > not recommend using it if you want to figure out the *real* numbers. Do
> > you see a different behavior with your tests?
> >
> > So based on these findings I use ab and parse the output to make a
> > different report than ab does.
> >
> > I didn't find the answer for your original answer in the LWP::UserAgent
> > manpage, but probably the right address is the libwww list:
> >
> > The latest version of this library is likely to be
> > available from CPAN as well as:
> >
> > http://www.linpro.no/lwp/
> >
> > The best place to discuss this code is on the
> > <li...@perl.org> mailing list.
> >
> > ______________________________________________________________________
> > Stas Bekman | JAm_pH -- Just Another mod_perl Hacker
> > http://stason.org/ | mod_perl Guide http://perl.apache.org/guide
> > mailto:stas@stason.org | http://perl.org http://stason.org/TULARC/
> > http://singlesheaven.com| http://perlmonth.com http://sourcegarden.org
> > ----------------------------------------------------------------------
>
>
______________________________________________________________________
Stas Bekman | JAm_pH -- Just Another mod_perl Hacker
http://stason.org/ | mod_perl Guide http://perl.apache.org/guide
mailto:stas@stason.org | http://perl.org http://stason.org/TULARC/
http://singlesheaven.com| http://perlmonth.com http://sourcegarden.org
----------------------------------------------------------------------
Re: [OT] LWP: Regression testing a mod_perl server
Posted by Adi <ad...@certsite.com>.
Cool, thanks for the quick response, Stas.
I don't seem to have LWP::Parallel. Looks like I need to upgrade to the
newest version of LWP. But yes, I did notice that ab gives much more
accurate results. I was using Time::HiRes to simply measure the time it
takes from request to response, and it gave widely varying results. Which
was why I wanted to do a loop so I could divide the total time taken by the
number of iterations.
I'm glad to hear you're already working on it. I'll definitely help out
with your regression suite, because I haven't gotten too far on mine (I just
started working on it today).
Thanks for the LWP list ref.. I'll post the question there.
-Adi
Stas Bekman wrote:
>
> First I work on a similar suite as well. I didn't want to announce it yet,
> before I get some things working. But since you've mentioned it here, here
> we go... Hope to get the first version out of the door in a few days.
>
> Regarding LWP, I've used a similar to ab script as described at
> http://perl.apache.org/guide/performance.html and I've found the results
> are very different if you the same test with ab and LWP::Parallel, so I'd
> not recommend using it if you want to figure out the *real* numbers. Do
> you see a different behavior with your tests?
>
> So based on these findings I use ab and parse the output to make a
> different report than ab does.
>
> I didn't find the answer for your original answer in the LWP::UserAgent
> manpage, but probably the right address is the libwww list:
>
> The latest version of this library is likely to be
> available from CPAN as well as:
>
> http://www.linpro.no/lwp/
>
> The best place to discuss this code is on the
> <li...@perl.org> mailing list.
>
> ______________________________________________________________________
> Stas Bekman | JAm_pH -- Just Another mod_perl Hacker
> http://stason.org/ | mod_perl Guide http://perl.apache.org/guide
> mailto:stas@stason.org | http://perl.org http://stason.org/TULARC/
> http://singlesheaven.com| http://perlmonth.com http://sourcegarden.org
> ----------------------------------------------------------------------
Re: [OT] LWP: Regression testing a mod_perl server
Posted by Stas Bekman <sb...@stason.org>.
> I am trying to put together a comprehensive regression suite for CGI based
> mod_perl web applications using LWP. It involves sequences of CGI calls
> with expected results after each consecutive request. But I noticed
> something about LWP: it caches requests, almost like it expects *only*
> static content. So if you loop over a given sequence of HTTP requests, only
> the first one actually gets sent to the server.
>
> eg:
>
> foreach (1..5) {
> $ua->request(GET ('http://localhost/uri1'));
> $ua->request(GET ('http://localhost/uri2'));
> $ua->request(GET ('http://localhost/uri3'));
> }
>
> # this code will only send 3 requests to the webserver, NOT 15
>
> Has anyone done similar work making a regression test using LWP ? Is there
> an easy way to turn off this caching behavior?
First I work on a similar suite as well. I didn't want to announce it yet,
before I get some things working. But since you've mentioned it here, here
we go... Hope to get the first version out of the door in a few days.
Regarding LWP, I've used a similar to ab script as described at
http://perl.apache.org/guide/performance.html and I've found the results
are very different if you the same test with ab and LWP::Parallel, so I'd
not recommend using it if you want to figure out the *real* numbers. Do
you see a different behavior with your tests?
So based on these findings I use ab and parse the output to make a
different report than ab does.
I didn't find the answer for your original answer in the LWP::UserAgent
manpage, but probably the right address is the libwww list:
The latest version of this library is likely to be
available from CPAN as well as:
http://www.linpro.no/lwp/
The best place to discuss this code is on the
<li...@perl.org> mailing list.
______________________________________________________________________
Stas Bekman | JAm_pH -- Just Another mod_perl Hacker
http://stason.org/ | mod_perl Guide http://perl.apache.org/guide
mailto:stas@stason.org | http://perl.org http://stason.org/TULARC/
http://singlesheaven.com| http://perlmonth.com http://sourcegarden.org
----------------------------------------------------------------------