perl-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stas Bekman <>
Subject Re: dual server setup with 2.0?
Date Tue, 24 Apr 2001 04:42:43 GMT
On Mon, 23 Apr 2001, Doug MacEachern wrote:

> On Tue, 24 Apr 2001, Stas Bekman wrote:
> > Hey folks,
> >
> > I thought about the mod_perl 2.0 performance setups. So, since now one
> > process can have pools with interpreter and without it, would it make
> > sense to kill the dual server setup as we endorse now (where mod_perl is
> > the backend) and have only one server, with different pools.
> >
> > So would something like this faster than the current schema?
> >
> > Threads Pool A : lots of plain Apache threads.
> > Threads Pool P : a few Perl interpreters.
> this is not the way it works.  there is one thread "pool", which apache
> manages, no Perl interpreters are associated with these threads.
> when a request comes in, mod_perl checks to see if it should handle
> it.  if so, selects an interpreter from the interpreter pool and the
> current thread will then use that interpreter.  for how long depends on
> PerlInterpScope, which is per-request by default.
> > With a dual server setup mod_proxy talks to the back-end server.
> > How this actually will work within one 2.0 server? Is there going to be
> > some kind mod_proxy like facility that will know to dispatch the requests?
> see above.
> i don't know if 2.0 will obsolete the need for dual server setup.  with
> the right config parameters, maybe.

OK, so how mod_perl can then tell apache to delegate the generated
response to some light thread to spoon feed the client? Or can you release
the Perl interpreter back into the pool, before all of the responce has
reached the client?

Stas Bekman              JAm_pH     --   Just Another mod_perl Hacker       mod_perl Guide

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message