pure fileless requests
Hello, I use PerlTransHandler to decide if I want to intercept and serve a request or not. When the request is served by my content handler there is no file that corresponds to it, directly or indirectly. I used to just not call $r->filename(...) at all. But under Gentoo's overtightened default config it fails with access denied. My best guess -- because it maps to '/' by default and that's denied. I had to resort to calling $r->filename($r->document_root.'/index.html) which does solve my problem, but looks terrible. And apache still translates it to index.html.en, provides finfo() and so -- none of which I need. So, my question is -- what is the best way to tell Apache that there is no file associated with this request, don't bother trying to stat() any files or run any rules. Thanx! Oh, and I use latest 2.x versions for development, but the same code runs in 1.x and 2.x mod_perl/apache setups. Andrew. -- Report problems: http://perl.apache.org/bugs/ Mail list info: http://perl.apache.org/maillist/modperl.html List etiquette: http://perl.apache.org/maillist/email-etiquette.html
Re: [mp2] Apache::DBI and Apache::Status
Scott Scecina wrote:
I haven't been able to find an answer to this, so I thought I'd throw it
out here:
Apache::DBI used to "plugin" to Apache::Status in order to display the DBI
connections. Is this not supported in mp2?
Looks like it wasn't ported to mp2. The following patch should do the trick.
I hope Ask integrates it.
--- /tmp/DBI.pm 2004-11-05 16:09:14.779192529 -0500
+++ /tmp/DBI.pm.new 2004-11-05 16:21:03.967049395 -0500
@@ -193,9 +193,18 @@
# prepare menu item for Apache::Status
+use mod_perl;
+use constant MP2 => ($mod_perl::VERSION >= 1.99);
+my $apache_status_avail = 0;
+$apache_status_avail = 1
+if !MP2 and $INC{'Apache.pm'} # is Apache.pm loaded?
+and Apache->can('module') # really?
+and Apache->module('Apache::Status'); # Apache::Status too?
+$apache_status_avail = 1
+if MP2 and require Apache::Module
+and Apache::Module::loaded('Apache::Status');
Apache::Status->menu_item(
-
'DBI' => 'DBI connections',
sub {
my($r, $q) = @_;
@@ -207,9 +216,7 @@
return [EMAIL PROTECTED];
}
-) if ($INC{'Apache.pm'} # is Apache.pm loaded?
- and Apache->can('module') # really?
- and Apache->module('Apache::Status')); # Apache::Status too?
+) if $apache_status_avail;
1;
--
__
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
--
Report problems: http://perl.apache.org/bugs/
Mail list info: http://perl.apache.org/maillist/modperl.html
List etiquette: http://perl.apache.org/maillist/email-etiquette.html
RE: Global Variables - What is in Memory?
This clears things up. Thanks for your help. Justin -Original Message- From: Malcolm J Harwood [mailto:[EMAIL PROTECTED] Sent: Friday, November 05, 2004 9:22 AM To: [EMAIL PROTECTED] Subject: Re: Global Variables - What is in Memory? On Friday 5 November 2004 11:50 am, Justin Luster wrote: > > Apache::Registry should mangle scripts to seperate names so that they > > don't conflict. However if you are using the same package name in each > > case, the last definition to be loaded will be used. Ie. if both > > one/MyAdmin.pl and two/MyAdmin.pl define admin:main(), whichever file > > loaded last will have it's definition used. (Check your error logs for > > "redefinition" warnings). > I'm still not totally clear on this. > So if I have two scripts: > http://www.mysite.com/one/MyAdmin.pl > and > http://www.mysite.com/two/MyAdmin.pl > are you saying that the latest one loaded will be used? So if > http://www.mysite.com/two/MyAdmin.pl was loaded last then > http://www.mysite.com/one/MyAdmin.pl would actually use the script > contained in two/MyAdmin.pl? No. > The package part though does make sense. So the > last "package" loaded stays in memory? If you load a file with a package definition in it, then load another file that defines a package with the same name (which is what it looks like you are doing) you get a conflict. If you define a function in both package definitions with the same name, then whichever one happens to be loaded last will override the previous definition. Usually this results in a "Redefinition of package::functioname " message in your error log. Normally Apache::Registry only loads packages that are requested using "use package" once (there's a package that will check for changes to package files and reload them if they change, but you have to explicitly use it). However, you have your package definition in your scripts, which is where the problem occurs. Apache::Registry will load each script when it is first encountered or when it changes, and if you are running multiprocess apache then each child will load your scripts in an arbitary order. So some processes will have the definition from one/MyAdmin.pl and some from two/MyAdmin.pl with no way of predicting which, or which process will answer a given request. With your scripts, you end up with the following packages being known to mod_perl: package Apache::ROOT::webdir::one::MyAdminpl package Apache::ROOT::webdir::two::MyAdminpl package mylib ... one/MyAdmin.pl contains the following package definitions: package Apache::ROOT::webdir::one::MyAdminpl package mylib two/MyAdmin.pl contains the following package definitions: package Apache::ROOT::webdir::two::MyAdminpl package mylib However the two definitions of "mylib" are different. You see the problem? -- "You heard Mr. Garibaldi. You did the right thing." "Darling, I did the necessary thing. That is not always the same as the right thing." - Janice and Laura Rosen in Babylon 5:"The Quality of Mercy" -- Report problems: http://perl.apache.org/bugs/ Mail list info: http://perl.apache.org/maillist/modperl.html List etiquette: http://perl.apache.org/maillist/email-etiquette.html -- Report problems: http://perl.apache.org/bugs/ Mail list info: http://perl.apache.org/maillist/modperl.html List etiquette: http://perl.apache.org/maillist/email-etiquette.html
RE: Global Variables - What is in Memory?
On Fri, 2004-11-05 at 12:18, Justin Luster wrote: > So what you are saying is that > > http://www.mysite.com/one/MyAdmin.pl > and > http://www.mysite.com/two/MyAdmin.pl > > get different package names (because they are in different directories) > and so stay separate, but because I have declared a package inside of > them with a common name that package goes into a common package name and > will be shared by both scripts? Yes. Symbols like subs and globals go into the current package. (Packages have no relation to file names, except that "use My::Foo" will try to guess a file name based on what you give it.) You've declared the same package in both places, so one is overwriting the other. The link that Malcolm gave provides a good description of what happens to your script when you run it under Registry. - Perrin -- Report problems: http://perl.apache.org/bugs/ Mail list info: http://perl.apache.org/maillist/modperl.html List etiquette: http://perl.apache.org/maillist/email-etiquette.html
Re: Global Variables - What is in Memory?
On Friday 5 November 2004 11:50 am, Justin Luster wrote: > > Apache::Registry should mangle scripts to seperate names so that they > > don't conflict. However if you are using the same package name in each > > case, the last definition to be loaded will be used. Ie. if both > > one/MyAdmin.pl and two/MyAdmin.pl define admin:main(), whichever file > > loaded last will have it's definition used. (Check your error logs for > > "redefinition" warnings). > I'm still not totally clear on this. > So if I have two scripts: > http://www.mysite.com/one/MyAdmin.pl > and > http://www.mysite.com/two/MyAdmin.pl > are you saying that the latest one loaded will be used? So if > http://www.mysite.com/two/MyAdmin.pl was loaded last then > http://www.mysite.com/one/MyAdmin.pl would actually use the script > contained in two/MyAdmin.pl? No. > The package part though does make sense. So the > last "package" loaded stays in memory? If you load a file with a package definition in it, then load another file that defines a package with the same name (which is what it looks like you are doing) you get a conflict. If you define a function in both package definitions with the same name, then whichever one happens to be loaded last will override the previous definition. Usually this results in a "Redefinition of package::functioname " message in your error log. Normally Apache::Registry only loads packages that are requested using "use package" once (there's a package that will check for changes to package files and reload them if they change, but you have to explicitly use it). However, you have your package definition in your scripts, which is where the problem occurs. Apache::Registry will load each script when it is first encountered or when it changes, and if you are running multiprocess apache then each child will load your scripts in an arbitary order. So some processes will have the definition from one/MyAdmin.pl and some from two/MyAdmin.pl with no way of predicting which, or which process will answer a given request. With your scripts, you end up with the following packages being known to mod_perl: package Apache::ROOT::webdir::one::MyAdminpl package Apache::ROOT::webdir::two::MyAdminpl package mylib ... one/MyAdmin.pl contains the following package definitions: package Apache::ROOT::webdir::one::MyAdminpl package mylib two/MyAdmin.pl contains the following package definitions: package Apache::ROOT::webdir::two::MyAdminpl package mylib However the two definitions of "mylib" are different. You see the problem? -- "You heard Mr. Garibaldi. You did the right thing." "Darling, I did the necessary thing. That is not always the same as the right thing." - Janice and Laura Rosen in Babylon 5:"The Quality of Mercy" -- Report problems: http://perl.apache.org/bugs/ Mail list info: http://perl.apache.org/maillist/modperl.html List etiquette: http://perl.apache.org/maillist/email-etiquette.html
RE: Global Variables - What is in Memory?
Thanks. So what you are saying is that http://www.mysite.com/one/MyAdmin.pl and http://www.mysite.com/two/MyAdmin.pl get different package names (because they are in different directories) and so stay separate, but because I have declared a package inside of them with a common name that package goes into a common package name and will be shared by both scripts? Thanks Perrin. -Original Message- From: Perrin Harkins [mailto:[EMAIL PROTECTED] Sent: Friday, November 05, 2004 9:03 AM To: Justin Luster Cc: [EMAIL PROTECTED] Subject: Re: Global Variables - What is in Memory? Hi Justin, > use strict; > package mylib; > # Globals: Set them below > > $mylib::strGlobalStudyName = ""; You could also just say "our $strGlobalStudyName;" here. > I call this initialize function every time the script runs. Anyway I > forgot and so one of my global variables was not initialized. I was > shocked when I discovered the value from the global variable from > another user stick around. That's what they do. Globals don't go out of scope, so they retain their values. > I know global variables are bad but sometimes you have to use them. > What are some good work-a-rounds? Using lexicals, or sometimes singleton objects with accessors. It kind of depends on why you thought you needed to use a global in the first place. > But my big question is, is it safe to use global variables at all? Am > I guaranteed that if I initialize them at the start of my script that > they won't be set by another process before I use them later in my > script??? You are guaranteed of that, because each mod_perl request is handled by a separate process in mod_perl 1. (Even if you use a threaded MPM in MP2 you are still safe unless you explicitly declare the variable shared between threads.) > Today I was surprised when it appeared that my main script was in > memory. Yes, that's what Apache::Registry does. This is described in the Apache::Registry documentation and other places in the mod_perl docs online. > Well, the script in directory "one" started to run the version in > directory "two". I would understand this behavior if their was a > "required" library file, but in this case there are two main files. > Does the package "admin" stay in memory even though I did not > "require" it? Yes. Apache::Registry turns it into a module and requires it. That's why Registry is fast. If you didn't declare a package name in that file, each one would get a different auto-generated package name and they would not overlap. The automatic package name is based on the URL. Look at the Registry code if you want to see how this works. It's short and pure perl. > I have been lead to believe that it does not stay in memory because > when I update MyAdmin.pl with new code I see the new behavior without > having to reset Apache. It just checks to see if the file has been updated and reloads it if it has. - Perrin -- Report problems: http://perl.apache.org/bugs/ Mail list info: http://perl.apache.org/maillist/modperl.html List etiquette: http://perl.apache.org/maillist/email-etiquette.html -- Report problems: http://perl.apache.org/bugs/ Mail list info: http://perl.apache.org/maillist/modperl.html List etiquette: http://perl.apache.org/maillist/email-etiquette.html
Re: Global Variables - What is in Memory?
Hi Justin, > use strict; > package mylib; > # Globals: Set them below > > $mylib::strGlobalStudyName = ""; You could also just say "our $strGlobalStudyName;" here. > I call this initialize function every time the script runs. Anyway I > forgot and so one of my global variables was not initialized. I was > shocked when I discovered the value from the global variable from > another user stick around. That's what they do. Globals don't go out of scope, so they retain their values. > I know global variables are bad but sometimes you have to use them. > What are some good work-a-rounds? Using lexicals, or sometimes singleton objects with accessors. It kind of depends on why you thought you needed to use a global in the first place. > But my big question is, is it safe to use global variables at all? Am > I guaranteed that if I initialize them at the start of my script that > they wonât be set by another process before I use them later in my > script??? You are guaranteed of that, because each mod_perl request is handled by a separate process in mod_perl 1. (Even if you use a threaded MPM in MP2 you are still safe unless you explicitly declare the variable shared between threads.) > Today I was surprised when it appeared that my main script was in > memory. Yes, that's what Apache::Registry does. This is described in the Apache::Registry documentation and other places in the mod_perl docs online. > Well, the script in directory âoneâ started to run the version in > directory âtwoâ. I would understand this behavior if their was a > ârequiredâ library file, but in this case there are two main files. > Does the package âadminâ stay in memory even though I did not > ârequireâ it? Yes. Apache::Registry turns it into a module and requires it. That's why Registry is fast. If you didn't declare a package name in that file, each one would get a different auto-generated package name and they would not overlap. The automatic package name is based on the URL. Look at the Registry code if you want to see how this works. It's short and pure perl. > I have been lead to believe that it does not stay in memory because > when I update MyAdmin.pl with new code I see the new behavior without > having to reset Apache. It just checks to see if the file has been updated and reloads it if it has. - Perrin -- Report problems: http://perl.apache.org/bugs/ Mail list info: http://perl.apache.org/maillist/modperl.html List etiquette: http://perl.apache.org/maillist/email-etiquette.html
RE: Global Variables - What is in Memory?
Thank you for the great response to my questions.
> Why initialise your data in a seperate function? Is there a reason for
not
> doing:
> use strict;
> package mylib;
> $strGlobalStudyName = "value"?
My $mylib::strGlobalStudyName is not a constant. Every process might
need to change it. So I need to make sure to initialize it on every
call.
> Apache::Registry should mangle scripts to seperate names so that they
> don't
> conflict. However if you are using the same package name in each case,
the
> last definition to be loaded will be used. Ie. if both
> one/MyAdmin.pl and two/MyAdmin.pl define admin:main(), whichever file
> loaded
> last will have it's definition used. (Check your error logs for
> "redefinition" warnings).
I'm still not totally clear on this.
So if I have two scripts:
http://www.mysite.com/one/MyAdmin.pl
and
http://www.mysite.com/two/MyAdmin.pl
are you saying that the latest one loaded will be used? So if
http://www.mysite.com/two/MyAdmin.pl was loaded last then
http://www.mysite.com/one/MyAdmin.pl would actually use the script
contained in two/MyAdmin.pl? Given my experience with Mod_Perl I find
that hard to believe. The package part though does make sense. So the
last "package" loaded stays in memory?
I'm still struggling with this area.
Thanks,
Justin
-Original Message-
From: Malcolm J Harwood [mailto:[EMAIL PROTECTED]
Sent: Friday, November 05, 2004 7:09 AM
To: [EMAIL PROTECTED]
Subject: Re: Global Variables - What is in Memory?
On Thursday 4 November 2004 09:53 pm, Justin Luster wrote:
> Problem #1:
> I forgot to initialize a global variable! Bummer. In my code I
declare
> global variables like this:
> use strict;
> package mylib;
> # Globals: Set them below
> $mylib::strGlobalStudyName = "";
> I then initialize them with this:
> sub InitializeMylibGlobals
> {
> $mylib::strGlobalStudyName = "";
> }
Why initialise your data in a seperate function? Is there a reason for
not
doing:
use strict;
package mylib;
$strGlobalStudyName = "value"?
and then just referencing it as "$mylib::strGlobalStudyName" when you
need it?
> I call this initialize function every time the script runs. Anyway I
> forgot and so one of my global variables was not initialized. I was
> shocked when I discovered the value from the global variable from
> another user stick around.
What else would you expect? It's a global, therefor unless you change
the
value it holds the last thing it was set to. You are running
Apache::Registry
and thus the code is not recompiled every time, so the values of globals
from
previous requests are retained (on a per process basis).
> What are some good work-a-rounds?
- If you want to go that far, you could write your own handler that
always
initialises your data.
- If the values are constants, define them as such.
- Use a globals object, have the new function do your initialisation,
your
script can't access it without explictly creating the object (so you
know
your scripts can't forget to call it) and it goes out of scope at the
end of
your script so the values are discarded.
> But my big question is, is it safe to use global variables at all? Am
I
> guaranteed that if I initialize them at the start of my script that
they
> won't be set by another process before I use them later in my
> script???
They can't be set in another process, each process has it's own copy. It
can
be set by another script being run by the same apache process however.
Scripts are atomic so you can't have a situation where one script
affects
another that is still running (unless you deliberately set up shared
memory).
> Problem #2:
> I realize that when a Perl library file is "required" in goes into
> memory. And then the next time that that library is needed it will
just
> be used from memory and not have to be read in again. I thought that
> only files that were "required" went into memory. Today I was
surprised
> when it appeared that my main script was in memory.
This is what mod_perl does.
> The main file that the web users hit in their URL is called
"MyAdmin.pl".
> This file looks like this:
> use strict;
> admin::main();
> package admin;
> sub main
> {
> }
> 1;
> This file is not required by any other script. On my machine I had
two
> different versions of "MyAdmin.pl". They were in different locations
> so:
> http://www.mysite.com/one/MyAdmin.pl
> and
> http://www.mysite.com/two/MyAdmin.pl
>
> Well, the script in directory "one" started to run the version in
> directory "two". I would understand this behavior if their was a
> "required" library file, but in this case there are two main files.
Apache::Registry should mangle scripts to seperate names so that they
don't
conflict. However if you are using the same package name in each case,
the
last definition to be loaded will be used. Ie. if both
one/MyAdmin.pl and two/MyAdmin.pl define admin:main(), whichever
Re: How to get a core dump
Many Thanks Stass and Glenn, I'll try all this anf will get back.. On dv, 2004-11-05 at 13:38, Marc Gracia wrote: > Hi everybody. > I have a problem on a production cluster with a somewhat big mod_perl > app, and I just cannot get any clue of what is happening. > > The problem is that the servers just exit with Segmentation fault > randomly. > The problem is rare, hapens 10/20 times each day in each of the 6 > frontends, which globaly processes about 1.000.000 daily hits. > The global stats show about 30 Internal server errors daily, I don't > know if a segfault can cause an Internal Server Error on the client (I > suppose not, if the server dies, cannot send the 505), but the numbers > don't match anyway. > > I think a coredump will help me understand why it segfaults, but I don't > know how to make apache dump a coredump, I've tried a lot of recipes > found on internet with any success. Making things more complicated, this > problems only happens on the production systems, (Suppose only on some > pages...) so I cannot reproduce it on my test system. > > So, my question is... There is any way to force apache to dump a > coredump file? I suppose I'm forgotting something but I really > desperate... > > A secondary question is, some of the servers transforms all UTF8 strings > with "garbage" when some Segfault happens (Seems like double-encoded > UTF8, the page shows 3 or 4 chars for every UTF8 char...). The only way > to solve this is reboot the machine completely. > Is that related to this same problem? Or is an obscure UTF8 perl/Apache > problem? > > Many thanks, > I'm using mod_perl 1.29 with apache 1.3.31. > My perl conf: > > Summary of my perl5 (revision 5.0 version 8 subversion 0) configuration: > Platform: > osname=linux, osvers=2.4.20-2.48smp, > archname=i386-linux-thread-multi > uname='linux str' > config_args='-des -Doptimize=-O2 -march=i386 -mcpu=i686 -g > -Dmyhostname=localhost [EMAIL PROTECTED] -Dcc=gcc -Dcf_by=Red > Hat, Inc. -Dinstallprefix=/usr -Dprefix=/usr -Darchname=i386-linux > -Dvendorprefix=/usr -Dsiteprefix=/usr > -Dotherlibdirs=/usr/lib/perl5/5.8.0 -Duseshrplib -Dusethreads > -Duseithreads -Duselargefiles -Dd_dosuid -Dd_semctl_semun -Di_db > -Ui_ndbm -Di_gdbm -Di_shadow -Di_syslog -Dman3ext=3pm -Duseperlio > -Dinstallusrbinperl -Ubincompat5005 -Uversiononly -Dpager=/usr/bin/less > -isr' > hint=recommended, useposix=true, d_sigaction=define > usethreads=define use5005threads=undef' > useithreads=define usemultiplicity= > useperlio= d_sfio=undef uselargefiles=define usesocks=undef > use64bitint=undef use64bitall=un uselongdouble= > usemymalloc=, bincompat5005=undef > Compiler: > cc='gcc', ccflags ='-D_REENTRANT -D_GNU_SOURCE -DTHREADS_HAVE_PIDS > -DDEBUGGING -fno-strict-aliasing -I/usr/local/include > -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/gdbm', > optimize='', > cppflags='-D_REENTRANT -D_GNU_SOURCE -DTHREADS_HAVE_PIDS -DDEBUGGING > -fno-strict-aliasing -I/usr/local/include -I/usr/include/gdbm' > ccversion='', gccversion='3.2.2 20030213 (Red Hat Linux 8.0 > 3.2.2-1)', gccosandvers='' > gccversion='3.2.2 200302' > intsize=e, longsize= , ptrsize=p, doublesize=8, byteorder=1234 > d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=12 > ivtype='long' > k', ivsize=4' > ivtype='long' > known_ext, nvtype='double' > o_nonbl', nvsize=, Off_t='', lseeksize=8 > alignbytes=4, prototype=define > Linker and Libraries: > ld='gcc' > l', ldflags =' -L/usr/local/lib' > ldf' > libpth=/usr/local/lib /lib /usr/lib > libs=-lnsl -lgdbm -ldb -ldl -lm -lpthread -lc -lcrypt -lutil > perllibs= > libc=/lib/libc-2.3.1.so, so=so, useshrplib=true, libperl=libper > gnulibc_version='2.3.1' > Dynamic Linking: > dlsrc=dl_dlopen.xs, dlext=so', d_dlsymun=undef, ccdlflags='-rdynamic > -Wl,-rpath,/usr/lib/perl5/5.8.0/i386-linux-thread-multi/CORE' > cccdlflags='-fPIC' > ccdlflags='-rdynamic -Wl,-rpath,/usr/lib/perl5', lddlflags='s > Unicode/Normalize XS/A' > > > Characteristics of this binary (from libperl): > Compile-time options: DEBUGGING MULTIPLICITY USE_ITHREADS > USE_LARGE_FILES PERL_IMPLICIT_CONTEXT > Locally applied patches: > MAINT18379 > Built under linux > Compiled at Feb 18 2003 22:19:53 > %ENV: > PERL5LIB="/usr/eBD/perl" > @INC: > /usr/eBD/perl/i386-linux-thread-multi > /usr/eBD/perl > /usr/lib/perl5/5.8.0/i386-linux-thread-multi > /usr/lib/perl5/5.8.0 > /usr/lib/perl5/site_perl/5.8.0/i386-linux-thread-multi > /usr/lib/perl5/site_perl/5.8.0 > /usr/lib/perl5/site_perl > /usr/lib/perl5/vendor_perl/5.8.0/i386-linux-thread-multi > /usr/lib/perl5/vendor_perl/5.8.0 > /usr/lib/perl5/vendor_perl > /usr/lib/perl5/5.8.0/i386-linux-thread-multi > /usr/lib/perl5/5.8.0 > > > > -- Marc Gracia <[EMAIL PROTECTED]> Oasys Soft -- Report problems: http://perl.apache.org/bugs/ Mail list info: http://perl.a
Re: How to get a core dump
On Fri, Nov 05, 2004 at 01:38:44PM +0100, Marc Gracia wrote:
> So, my question is... There is any way to force apache to dump a
> coredump file? I suppose I'm forgotting something but I really
> desperate...
On Linux, create a directory that is writable by the httpd user
mkdir /var/tmp/apache-cores
chown httpd /var/tmp/apache-cores
chmod 700 /var/tmp/apache-cores
Set the dump directory in httpd.conf:
CoreDumpDirectory /var/tmp/apache-cores
Enable core dumps in your shell and then start Apache
ulimit -c unlimited(for bash; different for csh)
/usr/local/apache/httpd
To test that you can get a dump, start up an httpd process,
send a SIGABRT, SIGBUS, or SIGSEGV to one of the _children_,
and check that it dumped. The error log should indicate
that a possible dump was saved to /var/tmp/apache-cores.
kill -11
('kill -l' (that's a lowercase letter L) to get a list
of signals and their numbers for your system)
Of course, this will all be a lot more useful to you if you
build Apache with debugging enabled. When building, set
OPTIM=-g before './configure' in the Apache source dir, e.g.
OPTIM=-g ./configure ...
or before 'perl Makefile.PL' in the mod_perl source dir, e.g.
OPTIM=-g perl Makefile.PL ...
depending on how you build Apache and mod_perl.
Cheers,
Glenn
--
Report problems: http://perl.apache.org/bugs/
Mail list info: http://perl.apache.org/maillist/modperl.html
List etiquette: http://perl.apache.org/maillist/email-etiquette.html
Re: How to get a core dump
Marc Gracia wrote: [...] So, my question is... There is any way to force apache to dump a coredump file? I suppose I'm forgotting something but I really desperate... Please read: http://perl.apache.org/docs/2.0/devel/debug/c.html#Getting_the_core_File_Dumped A secondary question is, some of the servers transforms all UTF8 strings with "garbage" when some Segfault happens (Seems like double-encoded UTF8, the page shows 3 or 4 chars for every UTF8 char...). The only way to solve this is reboot the machine completely. Is that related to this same problem? Or is an obscure UTF8 perl/Apache problem? Not sure what is the problem here. Any difference if you upgrade to perl 5.8.5? UTF8 support is constantly being improved by newer perl versions. -- __ 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 -- Report problems: http://perl.apache.org/bugs/ Mail list info: http://perl.apache.org/maillist/modperl.html List etiquette: http://perl.apache.org/maillist/email-etiquette.html
[mp2] Apache::DBI and Apache::Status
I haven't been able to find an answer to this, so I thought I'd throw it out here: Apache::DBI used to "plugin" to Apache::Status in order to display the DBI connections. Is this not supported in mp2? I have this in my httpd.conf: PerlModule Apache::Status PerlRequire conf/mpstartup.pl "conf/mpstartup.pl" contains (among other things) use Apache::DBI; Yet, /perl-status still just shows: Embedded Perl version v5.8.4 for Apache/2.0.52 process 4044, running since Fri Nov 5 10:48:23 2004 PerlRequire'd Files Perl Configuration Compiled Registry Scripts Environment ISA Tree Symbol Table Dump Loaded Modules Inheritance Tree I see this on both Win32 and Linux. Thanks! - Scott Scott Scecina -- Report problems: http://perl.apache.org/bugs/ Mail list info: http://perl.apache.org/maillist/modperl.html List etiquette: http://perl.apache.org/maillist/email-etiquette.html
Re: Global Variables - What is in Memory?
On Thursday 4 November 2004 09:53 pm, Justin Luster wrote:
> Problem #1:
> I forgot to initialize a global variable! Bummer. In my code I declare
> global variables like this:
> use strict;
> package mylib;
> # Globals: Set them below
> $mylib::strGlobalStudyName = "";
> I then initialize them with this:
> sub InitializeMylibGlobals
> {
> $mylib::strGlobalStudyName = "";
> }
Why initialise your data in a seperate function? Is there a reason for not
doing:
use strict;
package mylib;
$strGlobalStudyName = "value"?
and then just referencing it as "$mylib::strGlobalStudyName" when you need it?
> I call this initialize function every time the script runs. Anyway I
> forgot and so one of my global variables was not initialized. I was
> shocked when I discovered the value from the global variable from
> another user stick around.
What else would you expect? It's a global, therefor unless you change the
value it holds the last thing it was set to. You are running Apache::Registry
and thus the code is not recompiled every time, so the values of globals from
previous requests are retained (on a per process basis).
> What are some good work-a-rounds?
- If you want to go that far, you could write your own handler that always
initialises your data.
- If the values are constants, define them as such.
- Use a globals object, have the new function do your initialisation, your
script can't access it without explictly creating the object (so you know
your scripts can't forget to call it) and it goes out of scope at the end of
your script so the values are discarded.
> But my big question is, is it safe to use global variables at all? Am I
> guaranteed that if I initialize them at the start of my script that they
> won't be set by another process before I use them later in my
> script???
They can't be set in another process, each process has it's own copy. It can
be set by another script being run by the same apache process however.
Scripts are atomic so you can't have a situation where one script affects
another that is still running (unless you deliberately set up shared memory).
> Problem #2:
> I realize that when a Perl library file is "required" in goes into
> memory. And then the next time that that library is needed it will just
> be used from memory and not have to be read in again. I thought that
> only files that were "required" went into memory. Today I was surprised
> when it appeared that my main script was in memory.
This is what mod_perl does.
> The main file that the web users hit in their URL is called "MyAdmin.pl".
> This file looks like this:
> use strict;
> admin::main();
> package admin;
> sub main
> {
> }
> 1;
> This file is not required by any other script. On my machine I had two
> different versions of "MyAdmin.pl". They were in different locations
> so:
> http://www.mysite.com/one/MyAdmin.pl
> and
> http://www.mysite.com/two/MyAdmin.pl
>
> Well, the script in directory "one" started to run the version in
> directory "two". I would understand this behavior if their was a
> "required" library file, but in this case there are two main files.
Apache::Registry should mangle scripts to seperate names so that they don't
conflict. However if you are using the same package name in each case, the
last definition to be loaded will be used. Ie. if both
one/MyAdmin.pl and two/MyAdmin.pl define admin:main(), whichever file loaded
last will have it's definition used. (Check your error logs for
"redefinition" warnings).
> Does the package "admin" stay in memory even though I did not "require"
> it?
It appears to be defined within your script, packages are not independent of
the files they are in, so you've defined and used it, the script is not
discarded (because you're using Apache::Registry) so the package stays in
memory.
> I have been lead to believe that it does not stay in memory because
> when I update MyAdmin.pl with new code I see the new behavior without
> having to reset Apache.
Apache::Registry caches scripts and reloads them when they change. So when you
edit one of your scripts it will be reloaded, becoming the "latest defintion"
of admin::main() and overwriting the previous version.
see also: http://perl.apache.org/docs/1.0/guide/intro.html#Apache__Registry
--
"The Universe doesn't give you any points for doing things that are easy."
- Sheridan to Garibaldi in Babylon 5:"The Geometry of Shadows"
--
Report problems: http://perl.apache.org/bugs/
Mail list info: http://perl.apache.org/maillist/modperl.html
List etiquette: http://perl.apache.org/maillist/email-etiquette.html
How to get a core dump
Hi everybody. I have a problem on a production cluster with a somewhat big mod_perl app, and I just cannot get any clue of what is happening. The problem is that the servers just exit with Segmentation fault randomly. The problem is rare, hapens 10/20 times each day in each of the 6 frontends, which globaly processes about 1.000.000 daily hits. The global stats show about 30 Internal server errors daily, I don't know if a segfault can cause an Internal Server Error on the client (I suppose not, if the server dies, cannot send the 505), but the numbers don't match anyway. I think a coredump will help me understand why it segfaults, but I don't know how to make apache dump a coredump, I've tried a lot of recipes found on internet with any success. Making things more complicated, this problems only happens on the production systems, (Suppose only on some pages...) so I cannot reproduce it on my test system. So, my question is... There is any way to force apache to dump a coredump file? I suppose I'm forgotting something but I really desperate... A secondary question is, some of the servers transforms all UTF8 strings with "garbage" when some Segfault happens (Seems like double-encoded UTF8, the page shows 3 or 4 chars for every UTF8 char...). The only way to solve this is reboot the machine completely. Is that related to this same problem? Or is an obscure UTF8 perl/Apache problem? Many thanks, I'm using mod_perl 1.29 with apache 1.3.31. My perl conf: Summary of my perl5 (revision 5.0 version 8 subversion 0) configuration: Platform: osname=linux, osvers=2.4.20-2.48smp, archname=i386-linux-thread-multi uname='linux str' config_args='-des -Doptimize=-O2 -march=i386 -mcpu=i686 -g -Dmyhostname=localhost [EMAIL PROTECTED] -Dcc=gcc -Dcf_by=Red Hat, Inc. -Dinstallprefix=/usr -Dprefix=/usr -Darchname=i386-linux -Dvendorprefix=/usr -Dsiteprefix=/usr -Dotherlibdirs=/usr/lib/perl5/5.8.0 -Duseshrplib -Dusethreads -Duseithreads -Duselargefiles -Dd_dosuid -Dd_semctl_semun -Di_db -Ui_ndbm -Di_gdbm -Di_shadow -Di_syslog -Dman3ext=3pm -Duseperlio -Dinstallusrbinperl -Ubincompat5005 -Uversiononly -Dpager=/usr/bin/less -isr' hint=recommended, useposix=true, d_sigaction=define usethreads=define use5005threads=undef' useithreads=define usemultiplicity= useperlio= d_sfio=undef uselargefiles=define usesocks=undef use64bitint=undef use64bitall=un uselongdouble= usemymalloc=, bincompat5005=undef Compiler: cc='gcc', ccflags ='-D_REENTRANT -D_GNU_SOURCE -DTHREADS_HAVE_PIDS -DDEBUGGING -fno-strict-aliasing -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/gdbm', optimize='', cppflags='-D_REENTRANT -D_GNU_SOURCE -DTHREADS_HAVE_PIDS -DDEBUGGING -fno-strict-aliasing -I/usr/local/include -I/usr/include/gdbm' ccversion='', gccversion='3.2.2 20030213 (Red Hat Linux 8.0 3.2.2-1)', gccosandvers='' gccversion='3.2.2 200302' intsize=e, longsize= , ptrsize=p, doublesize=8, byteorder=1234 d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=12 ivtype='long' k', ivsize=4' ivtype='long' known_ext, nvtype='double' o_nonbl', nvsize=, Off_t='', lseeksize=8 alignbytes=4, prototype=define Linker and Libraries: ld='gcc' l', ldflags =' -L/usr/local/lib' ldf' libpth=/usr/local/lib /lib /usr/lib libs=-lnsl -lgdbm -ldb -ldl -lm -lpthread -lc -lcrypt -lutil perllibs= libc=/lib/libc-2.3.1.so, so=so, useshrplib=true, libperl=libper gnulibc_version='2.3.1' Dynamic Linking: dlsrc=dl_dlopen.xs, dlext=so', d_dlsymun=undef, ccdlflags='-rdynamic -Wl,-rpath,/usr/lib/perl5/5.8.0/i386-linux-thread-multi/CORE' cccdlflags='-fPIC' ccdlflags='-rdynamic -Wl,-rpath,/usr/lib/perl5', lddlflags='s Unicode/Normalize XS/A' Characteristics of this binary (from libperl): Compile-time options: DEBUGGING MULTIPLICITY USE_ITHREADS USE_LARGE_FILES PERL_IMPLICIT_CONTEXT Locally applied patches: MAINT18379 Built under linux Compiled at Feb 18 2003 22:19:53 %ENV: PERL5LIB="/usr/eBD/perl" @INC: /usr/eBD/perl/i386-linux-thread-multi /usr/eBD/perl /usr/lib/perl5/5.8.0/i386-linux-thread-multi /usr/lib/perl5/5.8.0 /usr/lib/perl5/site_perl/5.8.0/i386-linux-thread-multi /usr/lib/perl5/site_perl/5.8.0 /usr/lib/perl5/site_perl /usr/lib/perl5/vendor_perl/5.8.0/i386-linux-thread-multi /usr/lib/perl5/vendor_perl/5.8.0 /usr/lib/perl5/vendor_perl /usr/lib/perl5/5.8.0/i386-linux-thread-multi /usr/lib/perl5/5.8.0 -- Report problems: http://perl.apache.org/bugs/ Mail list info: http://perl.apache.org/maillist/modperl.html List etiquette: http://perl.apache.org/maillist/email-etiquette.html
Re: [mp2] API questions.
THanks.
All suggestions works well.
Stas Bekman wrote:
Ruslan U. Zakirov wrote:
Hello.
1) I can't find how to get string that represent initial request
location. I mean 'http://www.example.com/foo...'. Is it possible? Now,
I use next code as workaround:
my $location = $r->dir_config( 'ServerName' );
unless( $location ) {
$location = 'http://'. join ':', $r->get_server_name,
$r->get_server_port;
}
better:
$location = 'http://' . $r->construct_server;
http://perl.apache.org/docs/2.0/api/Apache/URI.html#C_construct_server_
but your code doesn't include the URI.
my $uri = $r->parsed_uri->unparse;
But really what you are after is: $r->construct_url();
http://perl.apache.org/docs/2.0/api/Apache/URI.html#C_construct_url_
2) Can I get name of the current mod_perl hook. My module works well
as PerlTransHandler or other that executes later, but should return OK
if it's translating handler and DECLINED in other cases. I wish to
write something like next code:
if( $r->foo_method eq 'PerlTransHandler' ) {
return OK;
} else {
return DECLINED
}
my $callback = ModPerl::Util::current_callback();
http://perl.apache.org/docs/2.0/api/ModPerl/Util.html#C_current_callback_
3) I have two configuration variables in httpd.conf:
PerlSetVar StorageRoot /mnt/local/storage/music/
PerlSetVar BaseLocation /download
Can I declare 'Alias BaseLocation StorageRoot' with mod_perl API?
Without this alias Apache can't find file and I don't want to force
user to input same strings twice.
There is no direct API (is there a C API for that? if there is we can
add the perl glue), but you can always push httpd.conf text with:
$r->add_config()
http://perl.apache.org/docs/2.0/api/Apache/RequestUtil.html#C_add_config_
or $s->add_config()
http://perl.apache.org/docs/2.0/api/Apache/ServerUtil.html#C_add_config_
--
Report problems: http://perl.apache.org/bugs/
Mail list info: http://perl.apache.org/maillist/modperl.html
List etiquette: http://perl.apache.org/maillist/email-etiquette.html
