I thought that it was a module compiled into Apache MPM. That you use
when compiling mp2.
My bad
Jay Scherrer
On Wed, 2005-04-27 at 16:43 +0200, Tom Schindl wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> Jay Scherrer schrieb:
> | According to Schwartz, when he last spoke on the s
Tom Schindl wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Stas Bekman schrieb:
| Tom Schindl wrote:
| [...]
|
|> Perl with compiled threads support (used or not used) is much slower and
|> when new threads are started the whole memory is copied over because
|> there's no copy-on-write logic
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Stas Bekman schrieb:
| Tom Schindl wrote:
| [...]
|
|> Perl with compiled threads support (used or not used) is much slower and
|> when new threads are started the whole memory is copied over because
|> there's no copy-on-write logic like there is for f
Tom Schindl wrote:
[...]
Perl with compiled threads support (used or not used) is much slower and
when new threads are started the whole memory is copied over because
there's no copy-on-write logic like there is for fork.
Actually there is COW in the recent perls, but it doesn't perform very well.
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Jay Scherrer schrieb:
| According to Schwartz, when he last spoke on the subject, He recommended
| using prefork with mp2. I guess he wasn't impressed enough with threads
| yet.
|
True if you can avoid it but as stated in another mail when runing on
win
According to Schwartz, when he last spoke on the subject, He recommended
using prefork with mp2. I guess he wasn't impressed enough with threads
yet.
Jay Scherrer
On Wed, 2005-04-27 at 08:39 +0200, Tom Schindl wrote:
> -BEGIN PGP SIGNED MESSAGE-
> Hash: SHA1
>
> Foo Ji-Haw schrieb:
> | H
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Foo Ji-Haw schrieb:
| Hello Tom,
|
| Just to clarify: are you saying that forking is better than threading on
| windows? Even if it is, I will have to throw in IPC to talk between the
| forks?
No. What I'm trying to say:
When using Mod-Perl2 on win32 yo
Hello Tom,
Just to clarify: are you saying that forking is better than threading on
windows? Even if it is, I will have to throw in IPC to talk between the
forks?
Tom Schindl wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Foo Ji-Haw schrieb:
| Hello Michael,
|
| Are you saying that Perl wi
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Foo Ji-Haw schrieb:
| Hello Michael,
|
| Are you saying that Perl with ithreads is slower, therefore
| implementation of threading on Perl is generally avoided?
|
| I write apps for both Windows and UNIX platforms. I thought that
| iThreads is the bette
Hello Michael,
Are you saying that Perl with ithreads is slower, therefore
implementation of threading on Perl is generally avoided?
I write apps for both Windows and UNIX platforms. I thought that
iThreads is the better alternative to forking, which I think is not well
implemented in Windows.
Michael Peters wrote:
Foo Ji-Haw wrote:
Hello Philip,
One of the biggest complaints I hear (and voice) is that the linux
distros ship a perl with ithreads. It's slower and almost noone wants or
needs it. It looks like FBSD is doing what the majority of people want
it to do, so I wouldn't knock it :
Michael Peters wrote:
Foo Ji-Haw wrote:
Hello Philip,
You are suggesting that FBSD 4.x is not easy to compile ithreads via
ports. I wonder why FBSD even on 5.x does not come with ithreads
precompiled (Linux does!). But I find it quite a deterent to use FBSD
for multiple apps in the future.
One of
Foo Ji-Haw wrote:
> Hello Philip,
>
> You are suggesting that FBSD 4.x is not easy to compile ithreads via
> ports. I wonder why FBSD even on 5.x does not come with ithreads
> precompiled (Linux does!). But I find it quite a deterent to use FBSD
> for multiple apps in the future.
One of the bigge
Foo Ji-Haw wrote:
Hello Philip,
Thanks for the good reply. Appreciate it. That has been bugging me for
the longest time.
You are suggesting that FBSD 4.x is not easy to compile ithreads via
ports. I wonder why FBSD even on 5.x does not come with ithreads
precompiled (Linux does!). But I find it
Foo Ji-Haw wrote:
Hello Stas,
I'd like to know more about your suggestion of 'running separate
mod_perl servers for each host'. Can you elaborate on that? I mean, do
you mean to run multiple copies of Apache?
That's correct, Foo Ji-Haw. See: http://modperlbook.org/html/ch12_01.html
In particular:
Hello Stas,
I'd like to know more about your suggestion of 'running separate
mod_perl servers for each host'. Can you elaborate on that? I mean, do
you mean to run multiple copies of Apache?
The question is whether you want to use it. Perl w/ithreads not only
slows things down and requires mor
Hello Philip,
Thanks for the good reply. Appreciate it. That has been bugging me for
the longest time.
You are suggesting that FBSD 4.x is not easy to compile ithreads via
ports. I wonder why FBSD even on 5.x does not come with ithreads
precompiled (Linux does!). But I find it quite a deterent
Foo Ji-Haw wrote:
Hello guys,
I know this is not exactly the best fit for this mailing list, but I am
trying to use FreeBSD (4.11 + 5.3) for mp2 work. Hope that qualifies it!
The problem is that the perl that comes with FBSD is not compiled with
ithreads. I read that to run different modperl app
On Apr 25, 2005, at 10:21 PM, Foo Ji-Haw wrote:
Hello guys,
I know this is not exactly the best fit for this mailing list, but I
am trying to use FreeBSD (4.11 + 5.3) for mp2 work. Hope that
qualifies it!
you'll want
cd /usr/ports/lang/perl5.8
make WITH_THREADS=yes
this will not build on amd64 ar
Hello guys,
I know this is not exactly the best fit for this mailing list, but I am
trying to use FreeBSD (4.11 + 5.3) for mp2 work. Hope that qualifies it!
The problem is that the perl that comes with FBSD is not compiled with
ithreads. I read that to run different modperl apps on virtual sites
20 matches
Mail list logo
