On Fri, 2003-10-17 at 13:07, Geoffrey Young wrote:
> simran wrote:
> > Hi All,
> >
> > I have a PerlTransHandler where i am doing something to the effect of:
> >
> > my $request = instance Apache::Request(shift);
> > $request->document_root("/home/testuser/www");
> > $request->pnotes
Hi --
Not sure if this a mod_perl problem or not, but I seek
any advice to how I might be able to avoid to solve it
using mod_perl.
On my site http://www.aaa.com I have a link to
https://www.bbb.com/redirect.
/redirect is handled by Redirect.pm, a mod_perl 2
module which sets a cookie and redire
simran wrote:
Hi All,
I have a PerlTransHandler where i am doing something to the effect of:
my $request = instance Apache::Request(shift);
$request->document_root("/home/testuser/www");
$request->pnotes("test_key" => "test_value");
warn "Set document root to /home/testuser/www
Hi All,
I have a PerlTransHandler where i am doing something to the effect of:
my $request = instance Apache::Request(shift);
$request->document_root("/home/testuser/www");
$request->pnotes("test_key" => "test_value");
warn "Set document root to /home/testuser/www\n";
Then later
1. Problem Description:
When I run make test I receive the following message
/usr/local/apache2/bin/httpd -d /usr/src/mod_perl-1.99_10/t -f
/usr/src/mod_perl-1.99_10/t/conf/httpd.conf -DAPACHE2 -DPERL_USEITHREADS
using Apache/2.0.47 (prefork MPM)
waiting for serv
On 16 Oct 2003 17:43:14 -0400
Perrin Harkins <[EMAIL PROTECTED]> wrote:
> On Thu, 2003-10-16 at 16:42, Simon Dassow wrote:
> This is probably because Apache::Session locks until it gets
> destroyed, and globals never get destroyed.
Doh... ok, now i know, thanks :)
> > My current solution is to st
On Thu, 2003-10-16 at 16:42, Simon Dassow wrote:
> I want to have %session to be shared in a module.
> With ``use vars qw(%session);'' it doesnt seem to work (Apache is blocking
> after a few requests).
This is probably because Apache::Session locks until it gets destroyed,
and globals never get d
Tim Smith wrote:
This is my setup:
RedHat 8 - Kernel 2.4.20-20.8
Apache 2.0.47
perl 5.8.0
mod_perl 1.99_10
When I run make test I receive the following message
/usr/local/apache2/bin/httpd -d /usr/src/mod_perl-1.99_10/t -f
/usr/src/mod_perl-1.99_10/t/conf/httpd.conf -DAPACHE2 -DPERL_USEITHREADS
u
Hi Pe(rlers|ople),
I want to have %session to be shared in a module.
With ``use vars qw(%session);'' it doesnt seem to work (Apache is blocking
after a few requests).
My current solution is to store the reference via
``$r->pnotes("SESSION_DATA"=>\%session);'' and save $r into $self so the other met
This is my setup:
RedHat 8 - Kernel 2.4.20-20.8
Apache 2.0.47
perl 5.8.0
mod_perl 1.99_10
When I run make test I receive the following message
/usr/local/apache2/bin/httpd -d /usr/src/mod_perl-1.99_10/t -f
/usr/src/mod_perl-1.99_10/t/conf/httpd.conf -DAPACHE2 -DPERL_USEITHREADS
using Apache/2.0.
Ray Zimmerman wrote:
At 5:10 PM +0100 10/16/03, Ged Haywood wrote:
Have you searched the archives? There was a thread talking about load
balancing earlier in the year, see
http://groups.yahoo.com/group/modperl/message/49334
for example which might help.
Thanks, I hadn't seen this thread. The p
At 5:10 PM +0100 10/16/03, Ged Haywood wrote:
Have you searched the archives? There was a thread talking about load
balancing earlier in the year, see
http://groups.yahoo.com/group/modperl/message/49334
for example which might help.
Thanks, I hadn't seen this thread. The pointer to the URL Rewrit
Scott Beuker wrote:
Hi,
With MP1 I was able to do the following in order to enable
Spreadsheet::WriteExcel to work properly:
tie *XLS, 'Apache';
binmode(XLS);
my $workbook = Spreadsheet::WriteExcel->new(\*XLS);
But with MP2 I get the error 'Can't locate object method "TIEHANDLE" via
package "Ap
Hi,
With MP1 I was able to do the following in order to enable
Spreadsheet::WriteExcel to work properly:
tie *XLS, 'Apache';
binmode(XLS);
my $workbook = Spreadsheet::WriteExcel->new(\*XLS);
But with MP2 I get the error 'Can't locate object method "TIEHANDLE" via
package "Apache"'. I have bee
Well, in parallel with my investigations our sysadmin was looking at
upgrading mod_perl itself. He rebuilt it as V1.29 (still with Perl
5.8.1) and all the problems went away! So, while we can no longer tell
anyone just what was wrong, we can at least recommend a way to escape
the swamp ...
--
Perrin Harkins wrote:
On Thu, 2003-10-16 at 11:15, Mike Norton wrote:
Yes I understand this however when I watch a particular process
using Apache::VMonitor the process carries on serving requests
even once it has exceeeded the size I have specified
What platform are you on? There may be a dif
Perrin Harkins wrote:
Geoffrey Young wrote:
so I see no harm in letting the masses get their feet wet with 2.0
now. we have lots of people installing it on lots of different
platforms, so getting past the initial installation hurdle shouldn't
be any more difficult than with 1.0. after that, t
Jeff Boes wrote:
Stas Bekman wrote:
Jeff Boes wrote:
We moved our webserver from a machine where mod_perl was built
under Perl 5.6.1, to a server where it was built under 5.8.1. Now,
Perl scripts run but produce no browser output!
You mean a simple script:
print "Content-type: text/plain\n\n"
Hi there,
On Thu, 16 Oct 2003, Ray Zimmerman wrote:
> We have a web app which consists of some very expensive dynamic
> pages. Network bandwidth isn't an issue, but web-server CPU load is.
> [snip]
> the front-end server proxies to in round-robin style. With our
> typical load this has solved t
Perrin Harkins wrote:
I recently gave some advice on perlmonks.org to a person who was
concerned about what his migration path would be if he develops for
mod_perl 1 right now. I told him he will be safe if he uses
Apache::Registry, since he should be able to run his A::R scripts under
ModPerl::Re
On Thu, 2003-10-16 at 11:38, Mike Norton wrote:
> Platform is RedHat Advanced Server 2.1,
Have you modified the page size? A::SL assumes 4K page sizes on Linux.
You can change it easilly in the linux_size_check sub.
Also, A::VM is being more precise about bytes. There are really 1024 KB
in 1
>What platform are you on? There may be a difference in the size
>calculations that the two modules are using. VMonitor uses GTop to get
>sizes, while SizeLimit uses various approaches depending on platform.
>On Linux it reads the /proc filesystem.
Platform is RedHat Advanced Server 2.1,
Mike
On Thu, 2003-10-16 at 11:15, Mike Norton wrote:
> Yes I understand this however when I watch a particular process
> using Apache::VMonitor the process carries on serving requests
> even once it has exceeeded the size I have specified
What platform are you on? There may be a difference in the siz
>Processes don't exit until they finish serving a request, so they will
>be larger for the time it takes to finish. What are you looking at to
>determine size?
Yes I understand this however when I watch a particular process using Apache::VMonitor
the process carries on serving requests even onc
On Thu, 2003-10-16 at 10:55, Mike Norton wrote:
> We have just put the Apache::SizeLimit in place to try and cap the
> size of the processes however we still seem to see processes that are
> larger than the limit we set any ideas on why this would be ?
Processes don't exit until they finish servin
We have just put the
Apache::SizeLimit in place to try and cap the size of the processes however we
still seem to see processes that are larger than the limit we set any ideas on
why this would be ?
startup.pl
use
Apache::SizeLimit;$Apache::SizeLimit::MAX_PROCESS_SIZE = 3; # in KB, s
There is a 'secure' flag for cookies to make sure
they are only transfered during a secure session.
I would assume it would have to be set initially
via a secure request too.
--
Hari
On Tue, Oct 14, 2003 at 06:35:55PM -0700, Tofu Optimist wrote:
> Many thanks Hari.
>
> Turns I was having an "ac
That's actually the reason I didn't recommend starting with mp2 to this
guy: the documentation. It's easy for me to look at things and say "oh,
this corresponds to the 1.0 docs here", but to a newbie that will sound
like insanity. There's a lot more documented than there used to be, and
I am
We have a web app which consists of some very expensive dynamic
pages. Network bandwidth isn't an issue, but web-server CPU load is.
We used to run a front-end webserver, back-end webserver and database
server on the same machine. Recently we tried moving the back-end
webserver to two separate
Geoffrey Young wrote:
so I see no harm in letting the masses get their feet wet with 2.0 now.
we have lots of people installing it on lots of different platforms, so
getting past the initial installation hurdle shouldn't be any more
difficult than with 1.0. after that, the main API hasn't chan
Jeff Boes wrote:
Stas Bekman wrote:
You are running under mod_perl 1.0, right? Can we see the output of
perl -V? Anything in the error_log?
Here's the -V output from the script running under mod_perl (thanks,
Rafael!):
Summary of my perl5 (revision 5.0 version 8 subversion 1) configuration:
Rafael Garcia-Suarez wrote:
Jeff Boes wrote:
And here's what perl -V produces from the command line (is there a way
to do an equivalent thing from within a mod_perl script?):
use Config qw/myconfig/;
print myconfig;
there's also Apache::Status (included in the mod_perl distribution),
Jeff Boes wrote:
> And here's what perl -V produces from the command line (is there a way
> to do an equivalent thing from within a mod_perl script?):
use Config qw/myconfig/;
print myconfig;
Stas Bekman wrote:
Jeff Boes wrote:
We moved our webserver from a machine where mod_perl was built
under Perl 5.6.1, to a server where it was built under 5.8.1. Now,
Perl scripts run but produce no browser output!
You mean a simple script:
print "Content-type: text/plain\n\n";
print "Hello";
It's kind of a tough call what to recommend. He's a mod_perl newbie, so
I don't feel comfortable suggesting he start with mp 2 at the moment.
Ultimately the best approach is probably to write for mp 1 now, and port
it in a few years with the help of compat, but Registry is a tempting
alternative
35 matches
Mail list logo
