Re: Dynamic VirtualHost's
Well, In our case, the server doesn't have any mod_perl apps. Only uses mod_perl (Or perl sections, i don't really know if it's the same) to dinamicaly generate the virtualhosts.
That apache2 acts as a reverse proxy to controls several httpd servers in the private lan with the mod_perl apps. It's a wonderfull option, if you need it. Works really well, and adds some more features as caching, warning pages while servers ar stopped or in maintenance And also the most important part, the internal servers always use high bandwith connections, so the session blocking time is lower.
El dv 04 de 03 del 2005 a les 15:37 +1100, en/na David J Radunz va escriure:
On Thu, 2005-03-03 at 11:44 +0100, Marc Gràcia wrote:
> Is not necesary to restart apache... A "graceful" do the job.
> I'm using this approach, and is really quick and not visible for end
> users. (I have 50 or so virtualhosts, so maybe in your case can be a
> little slower).
> Maybe some time after the "graceful" we have a little more load
> (apache is replacing processes an all this) but works like a charm.
> I think this is easily tested in your setup.
>
>
Strange - Ive always found that 'graceful' (althou a great option) plays
havoc with mod_perl applications.
> El dc 02 de 03 del 2005 a les 16:18 +0100, en/na Roman va escriure:
> > Hi all,
> > Is it possible to dynamically create VirtualHost's using mod_perl,
> > taking the value of each directive in a database ? If it is, how can I
> > do exactly please ?
> > Because actually, I'm doing things like :
> > -
> > push(@{ $VirtualHost{'*:80'} }, \%conf);
> > -
> > It works well, but with this method I must periodically restart Apache
> > and I think it uses as many memory as when we write
> > 's in the Apache .conf file, does it ?
> > (there will be maybe thousands of VirtualHost's)
> >
> > Thank you in advance,
> > Roman.
> >
> >
> --
> Marc Gràcia <[EMAIL PROTECTED]>
This email and any files transmitted with it are confidential and intended solely for the
use of the individual or entity to whom they are addressed. Please notify the sender
immediately by email if you have received this email by mistake and delete this email
from your system. Please note that any views or opinions presented in this email are solely
those of the author and do not necessarily represent those of the organisation.
Finally, the recipient should check this email and any attachments for the presence of
viruses. The organisation accepts no liability for any damage caused by any virus
transmitted by this email.
Re: [BUG] A few tests fail on RC4/5.6.1/Apache 1.3.26
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Stas Bekman wrote: | | Dominique, your report misses an essential info. Please review: | http://perl.apache.org/docs/2.0/user/help/help.html#_C_make_test___Failures | | | | Thanks. | Sorry (again). t/apache/content_length_header.1..27 # testing : GET /TestApache__content_length_header C-L header # expected: 0 # received: undef not ok 2 # testing : GET /TestApache__content_length_header?set_content_length C-L header # expected: 0 # received: 25 not ok 5 # testing : HEAD /TestApache__content_length_header?set_content_length C-L header # expected: undef # received: 25 not ok 17 I read t/logs/error_log and did not achieve enlightenment: END in modperl_extra.pl, pid=12592 [Fri Mar 04 11:33:55 2005] [notice] Apache/2.0.52 (Debian GNU/Linux) world domination series/2.0 mod_perl/1.999.21 Perl/v5.6.1 configured - -- resuming normal operations [Fri Mar 04 11:33:55 2005] [info] Server built: Nov 10 2004 11:57:35 [Fri Mar 04 11:33:55 2005] [debug] /build/buildd/apache2-2.0.52/build-tree/apache2/server/mpm/prefork/prefork.c(955): AcceptMutex: sysvsem (default: sysvsem) [Fri Mar 04 11:33:57 2005] [info] Child process pid=12600 is exiting END in modperl_extra.pl, pid=12600 [Fri Mar 04 11:33:57 2005] [info] Child process pid=12601 is exiting END in modperl_extra.pl, pid=12601 [Fri Mar 04 11:33:57 2005] [info] removed PID file /home/dom/ftp/mod_perl-2.0.0-RC4/t/logs/httpd.pid (pid=12596) [Fri Mar 04 11:33:57 2005] [notice] caught SIGTERM, shutting down END in modperl_extra.pl, pid=12596 - -- Dominique QUATRAVAUX Ingénieur senior 01 44 42 00 08 IDEALX -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.5 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFCKDp8MJAKAU3mjcsRAnMwAJ9uf9T736aQbWTLbaF5hxE13e6nBwCfR3vk xJK6jOmrNBxdsonCx3P5Tdk= =5vsv -END PGP SIGNATURE-
Re: Win32 + ModPerl 2 + Apache 2 + NPH script = confusing header situation
Formhandle wrote: Stas Bekman wrote: Please upgrade first to the latest mod_perl (2.0.0-RC4 or 1.999.21) and then try again: http://perl.apache.org/download/index.html http://perl.apache.org/docs/2.0/os/win32/install.html I tried but the PPM package was not available anywhere. I don't have the capacity to compile this stuff myself. I tried manually getting the latest from CPAN but it's all source and I kept running into compile/install issues. If you could perhaps point me to a binary version compatible with 5.6.x that is actually available somewhere for download, I would be most appreciative. has it to be perl 5.6.x? We are now at 5.8.6. You can get it from http://www.activestate.com/Products/ActivePerl/ and then you can get the ppms as described in the URL above. Please keep in mind I'm a self-taught programmer and, contrary to common perception, this mod_perl stuff is not the easiest thing in the world to install and keep in sync with Apache. I would love to be up-to-date with the latest and greatest if there were a simple and non-convoluted way to do so.
segmentation fault
Hi, I am using a very simple program which gets some data from a MySQL table, prints a form and that data. The server is: Apache/2.0.52 (Unix) PHP/5.0.2 mod_perl/1.99_19 Perl/v5.8.4 Server at 10.50.28.50 Port 80 Sometimes the program works fine, but very often it gives a 404 Not Found error. I have taken a look in the logs, and I have seen the following error, with no other errors that should tell me more: [Fri Mar 04 14:41:24 2005] [notice] child pid 4368 exit signal Segmentation fault (11) My program is not very big, and it uses the CGI and DBI modules directly, without other modules. Do you have any idea what could be the problem? I remember that when I have installed mod_perl, I couldn't install it using "make test" because it gave me some errors, so I have used just "make install". It gave me some errors that time that I couldn't understand at all and didn't know what to do to solve them... Thank you. Teddy
Verify mod_perl installation
Hi, Is there a way to test if mod_perl 2.0 was installed correctly? I have some problems with a program and I think mod_perl installation might be the source. Thank you. Teddy
Re: segmentation fault
On Fri, 4 Mar 2005 14:57:59 +0200, Octavian Rasnita <[EMAIL PROTECTED]> wrote: [snip] > I remember that when I have installed mod_perl, I couldn't install it using > "make test" because it gave me some errors, so I have used just "make > install". > It gave me some errors that time that I couldn't understand at all and > didn't know what to do to solve them... If make test failed then something is wrong. See the link and repost. http://perl.apache.org/docs/2.0/user/help/help.html#_C_make_test___Failures --Ian
Re: Win32 + ModPerl 2 + Apache 2 + NPH script = confusing header situation
On Fri, 4 Mar 2005, Tom Schindl wrote: > Formhandle wrote: > > Stas Bekman wrote: > > > >> Please upgrade first to the latest mod_perl (2.0.0-RC4 or 1.999.21) > >> and then try again: > >> http://perl.apache.org/download/index.html > >> http://perl.apache.org/docs/2.0/os/win32/install.html > > > > > > I tried but the PPM package was not available anywhere. > > I don't have the capacity to compile this stuff myself. > > I tried manually getting the latest from CPAN but it's > > all source and I kept running into compile/install > > issues. > > > > If you could perhaps point me to a binary version > > compatible with 5.6.x that is actually available > > somewhere for download, I would be most appreciative. > > has it to be perl 5.6.x? We are now at 5.8.6. You can get it from > http://www.activestate.com/Products/ActivePerl/ and then you can get the >ppms as described in the URL above. In fact, for mod_perl-2 (Apache/2.0.x), perl-5.8.x (ie, an ActivePerl 8xx build) is required for Win32; the mod_perl ppm package at http://theoryx5.uwinnipeg.ca/ppmpackages/ for ActivePerl 6xx works only with Apache/1.3. -- best regards, randy kobes
Re: segmentation fault
On Fri, 4 Mar 2005, Octavian Rasnita wrote: > Hi, > > I am using a very simple program which gets some data from a MySQL table, > prints a form and that data. > > The server is: > Apache/2.0.52 (Unix) PHP/5.0.2 mod_perl/1.99_19 Perl/v5.8.4 Server at > 10.50.28.50 Port 80 > > Sometimes the program works fine, but very often it gives a 404 Not Found > error. > > I have taken a look in the logs, and I have seen the following error, with > no other errors that should tell me more: > [Fri Mar 04 14:41:24 2005] [notice] child pid 4368 exit signal Segmentation > fault (11) > > My program is not very big, and it uses the CGI and DBI > modules directly, without other modules. > > Do you have any idea what could be the problem? > > I remember that when I have installed mod_perl, I couldn't > install it using "make test" because it gave me some > errors, so I have used just "make install". It gave me > some errors that time that I couldn't understand at all > and didn't know what to do to solve them... It's hard to tell what the problem might be without more details (version information of mod_perl, Apache, and Perl), as well as a minimal script illustrating the problem. However, it's probably best to first look at the "make test" problems, as they might be related. Such problems can be reported according to the guidelines at http://perl.apache.org/docs/2.0/user/help/help.html which will include information on your setup that will help in examining this. -- best regards, randy kobes
Re: Verify mod_perl installation
Hmm, the test-suite comming with mod_perl, download? Tom Octavian Rasnita wrote: Hi, Is there a way to test if mod_perl 2.0 was installed correctly? I have some problems with a program and I think mod_perl installation might be the source. Thank you. Teddy
Re: Win32 + ModPerl 2 + Apache 2 + NPH script = confusing header situation
Randy Kobes wrote: On Fri, 4 Mar 2005, Tom Schindl wrote: [...] has it to be perl 5.6.x? We are now at 5.8.6. You can get it from http://www.activestate.com/Products/ActivePerl/ and then you can get the ppms as described in the URL above. In fact, for mod_perl-2 (Apache/2.0.x), perl-5.8.x (ie, an ActivePerl 8xx build) is required for Win32; the mod_perl ppm package at http://theoryx5.uwinnipeg.ca/ppmpackages/ for ActivePerl 6xx works only with Apache/1.3. Thanks Randy for pointing this out. I'm not a win32 user so I didn't know that but next time I'll do so :-). Tom
Re: Win32 + ModPerl 2 + Apache 2 + NPH script = confusing header situation
Formhandle wrote: Stas Bekman wrote: Please upgrade first to the latest mod_perl (2.0.0-RC4 or 1.999.21) and then try again: http://perl.apache.org/download/index.html http://perl.apache.org/docs/2.0/os/win32/install.html [...] Please keep in mind I'm a self-taught programmer and, contrary to common perception, this mod_perl stuff is not the easiest thing in the world to install and keep in sync with Apache. Yes, but the newer version have the bug that you've reported fixed. I would love to be up-to-date with the latest and greatest if there were a simple and non-convoluted way to do so. Move away from win32 to a decent OS and you will have plenty to choose from. -- __ Stas BekmanJAm_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
Re: [BUG] A few tests fail on RC4/5.6.1/Apache 1.3.26
Dominique Quatravaux wrote: t/apache/content_length_header.1..27 # testing : GET /TestApache__content_length_header C-L header # expected: 0 # received: undef not ok 2 # testing : GET /TestApache__content_length_header?set_content_length C-L header # expected: 0 # received: 25 not ok 5 # testing : HEAD /TestApache__content_length_header?set_content_length C-L header # expected: undef # received: 25 not ok 17 Hmm, tried with 2.0.52 + perl 5.6.1 on linux, and it works fine. What's confusing is that your subject says Apache 1.3.26, whereas your report says Apache 2.0.52, whereas the error indicates that you really run Apache 2.1.x. Take a look at t/apache/content_length_header.t that's where you can see that httpd 2.1 has changed behavior. Dominique, did you build Apache by yourself? Is it possible that you have Apache 2.0.x with some special patches from 2.1.x applied to it? My guess is that this is the case. In which case there is nothing we can do about it, since you aren't running a proper Apache 2.0.x but some sort of hybrid. -- __ Stas BekmanJAm_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
Re: [BUG MP1/MP2] Apache should bail out in case of syntax errors in section
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Stas Bekman wrote:
| Unfortunately we can't have a test for this, since if the test
| suite doesn't start, we have a problem :( There should be a special
| small test suite for that I suppose.
I've cobbled the following together for this purpose, see below. It's
quite ugly but it appears to work.
| I know Philippe is moving ships at the moment, so it might be some
| time before he replies, but I'll put it on the TODO list, so it
| doesn't fall between chairs.
Maybe Philippe could factor out something like my
"Apache::TempTestConfig" package, only better? (IMHO
Apache::TestServer needs to be better decoupled from the rest of
Apache::Test in order to check cleanly for Apache startup failures).
Regards,
== t/badsyntax.t
use strict;
use warnings FATAL => 'all';
use Apache::Test;
use Apache::TestServer;
use Apache::TestConfigPerl;
use IO::Socket::INET;
{
~# This class is a copy-cat of "Apache::TestConfig" with no
~# persistence, better suited to running temporary instances of
~# Apache.
~package Apache::TempTestConfig;
~use IO::Socket::INET;
~use base 'Apache::TestConfig';
~sub new {
~my ($class, $uniquename) = @_;
~my %vars = %{Apache::TestConfig->thaw()->{vars}};
~delete $vars{thaw};
~my $self = $class->SUPER::new
~(%vars,
~ port => "select", # Get out of the way of other instances
~ t_logs => $vars{t_logs}."/$uniquename-httpd",
~ t_conf_file => $vars{t_conf}."/$uniquename-httpd.conf");
~$self->httpd_config(); # Can be done immediately, we know
~# everything we need.
~$self->configure_libmodperl();
~$self;
~}
~# Disable persistence.
~sub custom_config_first_time { die }
~sub custom_config_save { }
~# Disable the standard included files
~sub generate_extra_conf { [] }
~# Lifted from $self->server()->start(), adapted so as not to be as
~# picky when server fails starting ('cause in some cases that's
~# actually what we want to test)
~my %one_process = (1 => '-X', 2 => '-D ONE_PROCESS');
~sub server_start {
~my ($self, $timeout) = @_;
~$timeout ||= 5;
~my $server=$self->server();
~my $cmd = $server->start_cmd;
~my $vars = $self->{vars};
~my $httpd = $vars->{httpd} || 'unknown';
~if (Apache::TestConfig::WIN32) {
~#make sure only 1 process is started for win32
~#else Kill will only shutdown the parent
~my $one_process = $server->version_of(\%one_process);
~require Win32::Process;
~my $obj;
~# We need the "1" below to inherit the calling processes
~# handles when running Apache::TestSmoke so as to properly
~# dup STDOUT/STDERR
~Win32::Process::Create($obj,
~ $httpd,
~ "$cmd $one_process",
~ 1,
Win32::Process::NORMAL_PRIORITY_CLASS(),
~ '.');
~unless ($obj) {
~warn "Could not start the server: " .
~Win32::FormatMessage(Win32::GetLastError());
~return 0;
~}
~$self->{win32obj} = $obj;
~}
~defined(my $pid = fork) or die "Can't fork: $!";
~unless ($pid) { # child
~exec "$cmd" or do {
~my $status = $? >> 8;
~#error "httpd didn't start! $status";
~CORE::exit ($status);
~};
~}
~my $port = $self->select_first_port();
~for(my $i=0; $i < $timeout; $i++, sleep(1)) {
~return 1 if (IO::Socket::INET->new(PeerHost=>"localhost",
~ PeerPort=>$port));
~}
~return 0;
~}
~# Waits for the server to *really* be stopped.
~sub server_stop {
~my ($self) = @_;
~$self->server()->stop();
~sleep(1) while ($self->server()->ping() ||
~IO::Socket::INET->new
~(PeerHost=>"localhost",
~ PeerPort=>$self->select_first_port()));
~}
} # End of package Apache::TempTestConfig
plan tests => 2;
my $config = new Apache::TempTestConfig("runtimeerror");
$config->server_stop();
$config->postamble("
use Apache2;
die;
1;
");
$config->generate_httpd_conf;
$config->server_stop();
ok(! $config->server_start());
$config = new Apache::TempTestConfig("badsyntax");
$config->postamble("
use Apache2;
})] BAD SYNTAX })]
1;
");
$config->generate_httpd_conf;
sleep(2);
ok(! $config->server_start());
$config->server_stop();
- --
Dominique QUATRAVAUX Ingénieur senior
01 44 42 00 08 IDEALX
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.2.5 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigma
Re: [BUG] A few tests fail on RC4/5.6.1/Debian Apache 2.0.52-3
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Stas Bekman wrote: | Take a look at t/apache/content_length_header.t that's where you | can see that httpd 2.1 has changed behavior. True. Search'n-replacing of "have_min_apache_version(2.1)" with "1" cures the problem. | What's confusing is that your subject says Apache 1.3.26, Sorry yet again, wrong package (of course). And subject line changed. | whereas your report says Apache 2.0.52, whereas the error indicates | that you really run Apache 2.1.x. [...] Dominique, did you build | Apache by yourself? No indeed. This is apache2-mpm-prefork 2.0.52-3 from Debian testing, not the latest version in testing either. I cannot readily upgrade right now (apt-get says it's willing to install 2.0.53-5, but only if I give it permission to wipe out the rest of my system first :-) , but I'll try and recompile the .deb from its source and see if the problem goes away. | Is it possible that you have Apache 2.0.x with some special patches | from 2.1.x applied to it? My guess is that this is the case. In | which case there is nothing we can do about it, since you aren't | running a proper Apache 2.0.x but some sort of hybrid. | If Debian testing turns out to have this problem, I think that the test should at least be skipped on that distro, otherwise I probably won't be the last one to report this :-). Anyway, thanks a lot for your support. - -- Dominique QUATRAVAUX Ingénieur senior 01 44 42 00 08 IDEALX -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.5 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFCKK37MJAKAU3mjcsRAuEuAKCv4uoZcDa/COeBSW+UqncdbKoGyACgothQ T6rDiDs5fUC7MfWSESVyrF8= =Qnqh -END PGP SIGNATURE-
Re: [BUG] A few tests fail on RC4/5.6.1/Debian Apache 2.0.52-3
Dominique Quatravaux wrote: [...] | whereas your report says Apache 2.0.52, whereas the error indicates | that you really run Apache 2.1.x. [...] Dominique, did you build | Apache by yourself? No indeed. This is apache2-mpm-prefork 2.0.52-3 from Debian testing, not the latest version in testing either. I cannot readily upgrade right now (apt-get says it's willing to install 2.0.53-5, but only if I give it permission to wipe out the rest of my system first :-) , but I'll try and recompile the .deb from its source and see if the problem goes away. Most likely you will have the same problem with the newer debian package. | Is it possible that you have Apache 2.0.x with some special patches | from 2.1.x applied to it? My guess is that this is the case. In | which case there is nothing we can do about it, since you aren't | running a proper Apache 2.0.x but some sort of hybrid. | If Debian testing turns out to have this problem, I think that the test should at least be skipped on that distro, otherwise I probably won't be the last one to report this :-). No. 1) We can't skip the test just based on the fact that it's debian, since there will be other people who install a real Apache 2.0.x and not a mutant debian package, in which case it should not be skipped. 2) We can't please every distro out there. This is the same problem with distros adding future version patches to the perl-core and claiming to be what they aren't. If you want the mod_perl test suite to pass make sure that you have an authentic, unpatched version of Apache and of Perl. Otherwise you are on your own (well, you can ask support of your distro). -- __ Stas BekmanJAm_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
[Fwd: ApacheCon Europe 2005: Call for Presenters deadline extended!]
FYI Original Message Subject: ApacheCon Europe 2005: Call for Presenters deadline extended! Date: Fri, 4 Mar 2005 18:29:27 -0400 From: Lars Eilebrecht <[EMAIL PROTECTED]> Organization: The Apache Software Foundation To: [EMAIL PROTECTED] Call for Participation: ApacheCon Europe 2005 (CfP deadline extended!) == 18-22 July 2005, Stuttgart, Germany NEW SUBMISSION DEADLINE: Saturday, 12 March 2005, 22:30 EDT Come share your knowledge of Apache software at this educational and fun-filled gathering of Apache users, vendors, and friends. Apache Software Foundation members are designing the technical program for ApacheCon Europe 2005 that will include multiple tracks and dozens of technical sessions. We are particularly interested in session proposals covering: o Apache web server topics (installation, compilation, configuration, migration, ...) o All Apache Software Foundation projects (Jakarta, mod_perl, XML, SpamAssassin, et cetera) o scripting languages and dynamic content (Java, Perl, Python, TCL, XML, XSL, PHP, etc.) o Security and eCommerce o Performance tuning, load balancing, high availability o Technical and non-technical case studies o new technologies (web, Java, XML, anti-spam) Only educational sessions related to projects of the Apache Software Foundation or the Web in general will be considered (commercial sales or marketing presentation won't be accepted; please contact [EMAIL PROTECTED] if you're interested in giving a vendor presentation). Regards... -- Lars Eilebrecht - The Apache Software Foundation [EMAIL PROTECTED] -http://www.apachecon.com/ - To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
Re: [BUG] A few tests fail on RC4/5.6.1/Debian Apache 2.0.52-3
On Fri, 04 Mar 2005 15:07:36 -0500, Stas Bekman <[EMAIL PROTECTED]> wrote: > > | Is it possible that you have Apache 2.0.x with some special patches > > | from 2.1.x applied to it? My guess is that this is the case. In > > | which case there is nothing we can do about it, since you aren't > > | running a proper Apache 2.0.x but some sort of hybrid. > > | > > If Debian testing turns out to have this problem, I think that the > > test should at least be skipped on that distro, otherwise I probably > > won't be the last one to report this :-). > > No. > > 1) We can't skip the test just based on the fact that it's debian, since > there will be other people who install a real Apache 2.0.x and not a > mutant debian package, in which case it should not be skipped. This is caused by a patch that the Debian version applies. I've filed a bug report with them to remove it. http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=298143 The patch came from here: http://svn.apache.org/viewcvs.cgi/httpd-2.0/server/protocol.c?root=Apache&r1=1.150&r2=1.151 The patch was later reverted: http://cvs.apache.org/viewcvs.cgi?root=Apache-SVN&rev=110141&view=rev Dominique, you could also just remove that patch from your debian source package: debian/patches/035_HEAD_Content-Length_Fix_From_CVS And rebuild it. Contact me off-list if you need assistance with this before the Debian package is fixed.
Problem using Cleanuphandler for SizeLimit
Hello everyone, I just bump into something strange I wonder if anyone else here have experienced it or better have a fix. I have been using SizeLimit with FixupHandler since day one and it has been working wonderful, recently I noticed the documention have been changed now the CleanupHandler become the prefered choice of handler to handle SizeLimit. So I've picked a stage machine, make the change and let it run overnight, expecting nothing to go wrong and looking forward to make the change standard in subsequence deployment. I didn't have to wait very long, only after an hour or so, I start getting report that the box is acting 'funny', it starts rejecting requests and checking the log it seems some of the variables in the script being call, seems to be returning undefined. Restarted apache, it seems to work fine for a while then it comes again. I rolled back the change and a day has passed no more report of the problem have been reported since. I am running mod_perl 1.29 here, with apache 1.3.27 Many thanks. Tor. --
Re: Problem using Cleanuphandler for SizeLimit
On Fri, 2005-03-04 at 17:24 +0800, victor wrote: > I have been using SizeLimit with FixupHandler since day one and it has > been working wonderful, recently I noticed the documention have been > changed now the CleanupHandler become the prefered choice of handler to > handle SizeLimit. It still does the same thing it always did, i.e. if you use any phase other than cleanup it just pushes a cleanup handler to run at the end of the request. That seemed pointless to me, so I changed the documentation. > I didn't have to wait very long, only after an hour or so, I start > getting report that the box is acting 'funny', it starts rejecting > requests and checking the log it seems some of the variables in the > script being call, seems to be returning undefined. I can't think of any reason that change would make a difference. Do you have something else that might be trying to set a cleanup handler? - Perrin
Re: Win32 + ModPerl 2 + Apache 2 + NPH script = confusing header situation
On Fri, 4 Mar 2005, Stas Bekman wrote: > Formhandle wrote: > > Stas Bekman wrote: > > > >> Please upgrade first to the latest mod_perl (2.0.0-RC4 or 1.999.21) > >> and then try again: > >> http://perl.apache.org/download/index.html > >> http://perl.apache.org/docs/2.0/os/win32/install.html > [...] > > Please keep in mind I'm a self-taught programmer and, > > contrary to common perception, this mod_perl stuff is > > not the easiest thing in the world to install and keep > > in sync with Apache. > > Yes, but the newer version have the bug that you've reported fixed. > > > I would love to be up-to-date with the latest and > > greatest if there were a simple and non-convoluted way > > to do so. > > Move away from win32 to a decent OS and you will have > plenty to choose from. Not wanting to turn this into a discussion of the merits of different operating systems, but I'd suggest that, for binary packaging systems, ActiveState's ppm utility is pretty mature, and relatively straightforward to use. With any such system one can run into problems, but this isn't limited to win32 - witness the difficulties people have on linux with mod_perl rpm upgrades. Of course, on linux compiling things on your own is (often) a viable option, but I'd say the ppm system is sufficiently developed so as to make the lack of a compiler on Win32 not so much of an issue for an average user. As Stas pointed out at http://perl.apache.org/docs/2.0/os/win32/install.html, Win32 mod_perl-2 ppm packages are available. You should just be able to do, for ActivePerl 8xx, C:\> ppm install http://theoryx5.uwinnipeg.ca/ppms/mod_perl.ppd to install (this version is currently at the latest mod_perl-2 release candidate). If you have problems doing this, or if you find the docs lacking or confusing, let us know, so that we can improve things. -- best regards, randy
