Stas Bekman wrote:
Geoffrey Young wrote:
e.g.: % cvs co -D 20040601 -d mp2-20040601 modperl-2.0
yeah, I can try when I get the time (heh). but it might not be mod_perl's fault so I'll really need to do an httpd regression as well.
OK. Either way it's better to get the bugs fixed while they are hot :)
well, I still haven't been able to isolate this. howerver, I have noticed something interesting. with echo_timeout disabled, here is a record of my nightly builds:
Jun 10 03:02 /tmp/test-2.0.46-worker-perl-5.8.4.test.out Jun 10 03:34 /tmp/test-2.0.49-worker-perl-5.8.4.test.out Jun 10 05:24 /tmp/test-2.1-worker-maintperl.out Jun 10 06:06 /tmp/test-2.1-worker-perl-5.8.4.out Jun 10 06:26 /tmp/test-2.1-prefork-maintperl.out Jun 10 06:40 /tmp/test-2.1-prefork-perl-5.8.4.out Jun 10 07:22 /tmp/test-2.0-worker-perl-5.8.0.out Jun 10 08:01 /tmp/test-2.0-worker-maintperl.out Jun 10 08:39 /tmp/test-2.0-worker-perl-5.8.4.out
if you look at the times, you'll see that it takes about half an hour for the test suite to run, with the exception of 2.1 worker on maintperl, which is taking 2 hours. I have noticed this consistently over the past few nights.
maint is pretty close to 5.8.4, though I haven't updated my maint for quite some time. I might just do that soonish.
I don't know what the interaction is, but something is definitely odd in that combination performance-wise. just pointing it out for the record or in case something clicks in your head.
It's usually the ithreads test that runs the longest time, but it's really hard to tell by comparing the whole test suite. If you could do a similar analysis per test or at least a group of tests then it'll be much easier to tell.
-- __________________________________________________________________ Stas Bekman JAm_pH ------> Just Another mod_perl Hacker http://stason.org/ mod_perl Guide ---> http://perl.apache.org mailto:[EMAIL PROTECTED] http://use.perl.org http://apacheweek.com http://modperlbook.org http://apache.org http://ticketmaster.com
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
