mod_perl-2.0.4 with Apache 2.2.9 and perl 5.10.0 intermittent

mod_perl-2.0.4 with Apache 2.2.9 and perl 5.10.0 intermittent

am 27.01.2010 03:19:31 von Pas Argenio

--000e0cd48bee664bf2047e1c0819
Content-Type: text/plain; charset=ISO-8859-1

Sorry if this is naive, but has anyone gotten this combo to work? Upgraded
to Apache 2.2.14, now crash on almost every request, then next httpd picks
up the request and actually displays the page...

I've got massive core dumps all ending in malloc/free, truss shows last
operation was close()

--000e0cd48bee664bf2047e1c0819
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

Sorry if this is naive, but has anyone gotten this combo to work?=A0 Upgrad=
ed to Apache 2.2.14, now crash on almost every request, then next httpd pic=
ks up the request and actually displays the page...

I've got mas=
sive core dumps all ending in malloc/free, truss shows last operation was c=
lose()


--000e0cd48bee664bf2047e1c0819--

Re: mod_perl-2.0.4 with Apache 2.2.9 and perl 5.10.0 intermittent

am 27.01.2010 07:12:51 von Fred Moyer

What OS are you running? What is the output of 'perl -V'?

On Tue, Jan 26, 2010 at 6:19 PM, Pas Argenio wrote:
> Sorry if this is naive, but has anyone gotten this combo to work?=A0 Upgr=
aded
> to Apache 2.2.14, now crash on almost every request, then next httpd pick=
s
> up the request and actually displays the page...
>
> I've got massive core dumps all ending in malloc/free, truss shows last
> operation was close()
>

Re: mod_perl-2.0.4 with Apache 2.2.9 and perl 5.10.0 intermittent crashing

am 27.01.2010 17:42:54 von Tosh Cooey

I'm seeing the same thing on my DEV server which is:

Server: Apache/2.2.11 (Ubuntu) mod_perl/2.0.4 Perl/v5.10.0

I see it in the mornings when it has been sitting around all night doing
nothing, and the first couple hits result in core dumps/seg faults, then
after pounding REFRESH like a mad monkey a couple times it then works.

This is on an Amazon EC2 Ubuntu 9.04 jaunty AMI built by Eric Hammond
http://alestic.com http://ec2ubuntu-group.notlong.com

perl -V:

Summary of my perl5 (revision 5 version 10 subversion 0) configuration:
Platform:
osname=linux, osvers=2.6.24-23-server,
archname=i486-linux-gnu-thread-multi
uname='linux rothera 2.6.24-23-server #1 smp wed apr 1 22:22:14 utc
2009 i686 gnulinux '
config_args='-Dusethreads -Duselargefiles -Dccflags=-DDEBIAN
-Dcccdlflags=-fPIC -Darchname=i486-linux-gnu -Dprefix=/usr
-Dprivlib=/usr/share/perl/5.10 -Darchlib=/usr/lib/perl/5.10
-Dvendorprefix=/usr -Dvendorlib=/usr/share/perl5
-Dvendorarch=/usr/lib/perl5 -Dsiteprefix=/usr/local
-Dsitelib=/usr/local/share/perl/5.10.0
-Dsitearch=/usr/local/lib/perl/5.10.0 -Dman1dir=/usr/share/man/man1
-Dman3dir=/usr/share/man/man3 -Dsiteman1dir=/usr/local/man/man1
-Dsiteman3dir=/usr/local/man/man3 -Dman1ext=1 -Dman3ext=3perl
-Dpager=/usr/bin/sensible-pager -Uafs -Ud_csh -Ud_ualarm -Uusesfio
-Uusenm -DDEBUGGING=-g -Doptimize=-O2 -Duseshrplib
-Dlibperl=libperl.so.5.10.0 -Dd_dosuid -des'
hint=recommended, useposix=true, d_sigaction=define
useithreads=define, usemultiplicity=define
useperlio=define, d_sfio=undef, uselargefiles=define, usesocks=undef
use64bitint=undef, use64bitall=undef, uselongdouble=undef
usemymalloc=n, bincompat5005=undef
Compiler:
cc='cc', ccflags ='-D_REENTRANT -D_GNU_SOURCE -DDEBIAN
-fno-strict-aliasing -pipe -I/usr/local/include -D_LARGEFILE_SOURCE
-D_FILE_OFFSET_BITS=64',
optimize='-O2 -g',
cppflags='-D_REENTRANT -D_GNU_SOURCE -DDEBIAN -fno-strict-aliasing
-pipe -I/usr/local/include'
ccversion='', gccversion='4.3.3', gccosandvers=''
intsize=4, longsize=4, ptrsize=4, doublesize=8, byteorder=1234
d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=12
ivtype='long', ivsize=4, nvtype='double', nvsize=8, Off_t='off_t',
lseeksize=8
alignbytes=4, prototype=define
Linker and Libraries:
ld='cc', ldflags =' -L/usr/local/lib'
libpth=/usr/local/lib /lib /usr/lib /usr/lib64
libs=-lgdbm -lgdbm_compat -ldb -ldl -lm -lpthread -lc -lcrypt
perllibs=-ldl -lm -lpthread -lc -lcrypt
libc=/lib/libc-2.9.so, so=so, useshrplib=true,
libperl=libperl.so.5.10.0
gnulibc_version='2.9'
Dynamic Linking:
dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags='-Wl,-E'
cccdlflags='-fPIC', lddlflags='-shared -O2 -g -L/usr/local/lib'


Characteristics of this binary (from libperl):
Compile-time options: MULTIPLICITY PERL_DONT_CREATE_GVSV
PERL_IMPLICIT_CONTEXT PERL_MALLOC_WRAP USE_ITHREADS
USE_LARGE_FILES USE_PERLIO USE_REENTRANT_API
Built under linux
Compiled at Jun 26 2009 18:23:00
@INC:
/etc/perl
/usr/local/lib/perl/5.10.0
/usr/local/share/perl/5.10.0
/usr/lib/perl5
/usr/share/perl5
/usr/lib/perl/5.10
/usr/share/perl/5.10
/usr/local/lib/site_perl
.


Of course this could also just be all the illegitimate bastard apache
children I have been spawning being totally unprepared to answer a real
web request, but I'm keeping an eye on this server...

Tosh


Pas Argenio wrote:
> Sorry if this is naive, but has anyone gotten this combo to work?
> Upgraded to Apache 2.2.14, now crash on almost every request, then next
> httpd picks up the request and actually displays the page...
>
> I've got massive core dumps all ending in malloc/free, truss shows last
> operation was close()

--
McIntosh Cooey - Twelve Hundred Group LLC - http://www.1200group.com/

Re: mod_perl-2.0.4 with Apache 2.2.9 and perl 5.10.0 intermittent

am 27.01.2010 18:28:21 von Pas Argenio

--000e0cd5cb3a991f80047e28ba34
Content-Type: text/plain; charset=ISO-8859-1

I just fell back to Apache-2.2.8 but same problem. By the way, Tosh, how
did you dump the libperl info?
I'm getting more core dumps than you, but pages are displayed by the next
child to pick it up. Here is the most common stack trace (via Solaris'
pstack command):

Wed Jan 27 09:06:35 EST 2010
core '/tmp/core' of 13862: /usr/local/apache/bin/httpd -k start
fee56ec8 t_delete (528c90, 228, 0, fecb2e24, fef303a8, c20360) + 74
fee56ac4 realfree (528a60, 229, d995c, fee56eb0, 0, 528a58) + 8c
fee56610 _malloc_unlocked (528598, 6f8, 528590, ff, 0, 0) + 260
fee56394 malloc (4c8, 1, da058, fee563a0, fef303a8, fef3a518) + 4c
fec8ae8c Perl_safesysmalloc (4c8, 0, 2000000, fffffff8, c000004, c329d9) +
10
fecc2294 Perl_sv_grow (d17680, 4c8, 0, 2000000, 4cb, 0) + f8
fecbd094 Perl_sv_setsv_flags (84408, 4c7, 12c04540, d17680, d177e0,
ff800000) + ea0
fecbd5ac Perl_sv_mortalcopy (d177e0, 5e65d0, fed76000, 78, fed74400,
d17680) + 60
fecb2e24 Perl_pp_aassign (fed76294, 5b8c20, 5b8c0c, 5b8c1c, 0, fed74400) +
154
fecaa830 Perl_runops_standard (0, fecb2cd0, fed75c88, 91c2c0, fed75c00,
91cec0) + 1c
feca4ce0 Perl_call_sv (fed75c00, fed76000, fed76000, fed75c00, fed75c00,
91c2c0) + 4a4
fec1ca94 modperl_callback (0, 44b410, 44b450, ab4b8, 160080, fed76000) +
338
fec1d05c modperl_callback_run_handlers (fed75dec, 4, 44b450, 121dc8, ab4b8,
0) + 31c
fec1d32c modperl_callback_per_dir (6, 44b450, 1, 1, ffbff03c, 214678) + 24
fec19ab4 modperl_response_handler_run (44b450, 0, 1460, 0, fed761d8,
fed76000) + 18
fec19cc4 modperl_response_handler_cgi (44b450, 4484b8, 121520, 0, 99070, 9)
+ 134
0003bfe4 ap_run_handler (44b450, 0, c, 449018, 448728, 0) + 3c
0003c460 ap_invoke_handler (44b450, 77400, 44b450, 0, fe780020, 0) + b8
000547f8 ap_process_request (44b450, 0, 4, 44b450, 0, 443728) + 160
00051964 ap_process_http_connection (4436c8, 443430, 443430, 0, 998d0, 2) +
10c
00042790 ap_run_process_connection (4436c8, 443430, 443430, 0, 441438,
4453f8) + 3c
00067a10 child_main (0, 0, 99c00, 99800, 11177, 99800) + 42c
00067c74 make_child (67400, 0, 0, fe781020, 440fc8, 0) + ec
00067d44 startup_children (5, feebd280, 0, 10, 1cf4, a) + 68
000686b0 ap_mpm_run (a4878, 99800, ab4b8, 99c00, 99800, 99800) + 950
000291f4 main (a4878, 96c00, 99000, 99000, a2870, 0) + 780
00028594 _start (0, 0, 0, 0, 0, 0) + 5c

My MPM is straight pre-fork, default 5 children to start with.
Unfortunately, this stuff needs to go into production soon.



On Wed, Jan 27, 2010 at 11:42 AM, Tosh Cooey wrote:

> I'm seeing the same thing on my DEV server which is:
>
> Server: Apache/2.2.11 (Ubuntu) mod_perl/2.0.4 Perl/v5.10.0
>
> I see it in the mornings when it has been sitting around all night doing
> nothing, and the first couple hits result in core dumps/seg faults, then
> after pounding REFRESH like a mad monkey a couple times it then works.
>
> This is on an Amazon EC2 Ubuntu 9.04 jaunty AMI built by Eric Hammond
> http://alestic.com http://ec2ubuntu-group.notlong.com
>
> perl -V:
>
>
> Summary of my perl5 (revision 5 version 10 subversion 0) configuration:
> Platform:
> osname=linux, osvers=2.6.24-23-server,
> archname=i486-linux-gnu-thread-multi
> uname='linux rothera 2.6.24-23-server #1 smp wed apr 1 22:22:14 utc 2009
> i686 gnulinux '
> config_args='-Dusethreads -Duselargefiles -Dccflags=-DDEBIAN
> -Dcccdlflags=-fPIC -Darchname=i486-linux-gnu -Dprefix=/usr
> -Dprivlib=/usr/share/perl/5.10 -Darchlib=/usr/lib/perl/5.10
> -Dvendorprefix=/usr -Dvendorlib=/usr/share/perl5 -Dvendorarch=/usr/lib/perl5
> -Dsiteprefix=/usr/local -Dsitelib=/usr/local/share/perl/5.10.0
> -Dsitearch=/usr/local/lib/perl/5.10.0 -Dman1dir=/usr/share/man/man1
> -Dman3dir=/usr/share/man/man3 -Dsiteman1dir=/usr/local/man/man1
> -Dsiteman3dir=/usr/local/man/man3 -Dman1ext=1 -Dman3ext=3perl
> -Dpager=/usr/bin/sensible-pager -Uafs -Ud_csh -Ud_ualarm -Uusesfio -Uusenm
> -DDEBUGGING=-g -Doptimize=-O2 -Duseshrplib -Dlibperl=libperl.so.5.10.0
> -Dd_dosuid -des'
>
> hint=recommended, useposix=true, d_sigaction=define
> useithreads=define, usemultiplicity=define
>
> useperlio=define, d_sfio=undef, uselargefiles=define, usesocks=undef
> use64bitint=undef, use64bitall=undef, uselongdouble=undef
> usemymalloc=n, bincompat5005=undef
> Compiler:
> cc='cc', ccflags ='-D_REENTRANT -D_GNU_SOURCE -DDEBIAN
> -fno-strict-aliasing -pipe -I/usr/local/include -D_LARGEFILE_SOURCE
> -D_FILE_OFFSET_BITS=64',
> optimize='-O2 -g',
> cppflags='-D_REENTRANT -D_GNU_SOURCE -DDEBIAN -fno-strict-aliasing -pipe
> -I/usr/local/include'
> ccversion='', gccversion='4.3.3', gccosandvers=''
> intsize=4, longsize=4, ptrsize=4, doublesize=8, byteorder=1234
> d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=12
>
> ivtype='long', ivsize=4, nvtype='double', nvsize=8, Off_t='off_t',
> lseeksize=8
> alignbytes=4, prototype=define
> Linker and Libraries:
> ld='cc', ldflags =' -L/usr/local/lib'
> libpth=/usr/local/lib /lib /usr/lib /usr/lib64
> libs=-lgdbm -lgdbm_compat -ldb -ldl -lm -lpthread -lc -lcrypt
> perllibs=-ldl -lm -lpthread -lc -lcrypt
> libc=/lib/libc-2.9.so, so=so, useshrplib=true,
> libperl=libperl.so.5.10.0
> gnulibc_version='2.9'
> Dynamic Linking:
> dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags='-Wl,-E'
> cccdlflags='-fPIC', lddlflags='-shared -O2 -g -L/usr/local/lib'
>
>
>
> Characteristics of this binary (from libperl):
> Compile-time options: MULTIPLICITY PERL_DONT_CREATE_GVSV
> PERL_IMPLICIT_CONTEXT PERL_MALLOC_WRAP USE_ITHREADS
> USE_LARGE_FILES USE_PERLIO USE_REENTRANT_API
> Built under linux
> Compiled at Jun 26 2009 18:23:00
> @INC:
> /etc/perl
> /usr/local/lib/perl/5.10.0
> /usr/local/share/perl/5.10.0
> /usr/lib/perl5
> /usr/share/perl5
> /usr/lib/perl/5.10
> /usr/share/perl/5.10
> /usr/local/lib/site_perl
> .
>
>
> Of course this could also just be all the illegitimate bastard apache
> children I have been spawning being totally unprepared to answer a real web
> request, but I'm keeping an eye on this server...
>
> Tosh
>
>
>
> Pas Argenio wrote:
>
>> Sorry if this is naive, but has anyone gotten this combo to work?
>> Upgraded to Apache 2.2.14, now crash on almost every request, then next
>> httpd picks up the request and actually displays the page...
>>
>> I've got massive core dumps all ending in malloc/free, truss shows last
>> operation was close()
>>
>
> --
> McIntosh Cooey - Twelve Hundred Group LLC - http://www.1200group.com/
>

--000e0cd5cb3a991f80047e28ba34
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

I just fell back to Apache-2.2.8 but same problem.=A0 By the way, Tosh, how=
did you dump the libperl info?
I'm getting more core dumps than you=
, but pages are displayed by the next child to pick it up.=A0 Here is the m=
ost common stack trace (via Solaris' pstack command):


Wed Jan 27 09:06:35 EST 2010
core '/tmp/core' of 13862:=A0=
   /usr/local/apache/bin/httpd -k start
=A0fee56ec8 t_delete (528c90=
, 228, 0, fecb2e24, fef303a8, c20360) + 74
=A0fee56ac4 realfree (528a60,=
229, d995c, fee56eb0, 0, 528a58) + 8c

=A0fee56610 _malloc_unlocked (528598, 6f8, 528590, ff, 0, 0) + 260
=A0fe=
e56394 malloc   (4c8, 1, da058, fee563a0, fef303a8, fef3a518) + 4c
=
=A0fec8ae8c Perl_safesysmalloc (4c8, 0, 2000000, fffffff8, c000004, c329d9)=
+ 10

=A0fecc2294 Perl_sv_grow (d17680, 4c8, 0, 2000000, 4cb, 0) + f8
=A0fecbd=
094 Perl_sv_setsv_flags (84408, 4c7, 12c04540, d17680, d177e0, ff800000) + =
ea0
=A0fecbd5ac Perl_sv_mortalcopy (d177e0, 5e65d0, fed76000, 78, fed744=
00, d17680) + 60

=A0fecb2e24 Perl_pp_aassign (fed76294, 5b8c20, 5b8c0c, 5b8c1c, 0, fed74400)=
+ 154
=A0fecaa830 Perl_runops_standard (0, fecb2cd0, fed75c88, 91c2c0, =
fed75c00, 91cec0) + 1c
=A0feca4ce0 Perl_call_sv (fed75c00, fed76000, fed=
76000, fed75c00, fed75c00, 91c2c0) + 4a4

=A0fec1ca94 modperl_callback (0, 44b410, 44b450, ab4b8, 160080, fed76000) +=
338
=A0fec1d05c modperl_callback_run_handlers (fed75dec, 4, 44b450, 121=
dc8, ab4b8, 0) + 31c
=A0fec1d32c modperl_callback_per_dir (6, 44b450, 1,=
1, ffbff03c, 214678) + 24

=A0fec19ab4 modperl_response_handler_run (44b450, 0, 1460, 0, fed761d8, fed=
76000) + 18
=A0fec19cc4 modperl_response_handler_cgi (44b450, 4484b8, 12=
1520, 0, 99070, 9) + 134
=A00003bfe4 ap_run_handler (44b450, 0, c, 44901=
8, 448728, 0) + 3c

=A00003c460 ap_invoke_handler (44b450, 77400, 44b450, 0, fe780020, 0) + b8<=
br>=A0000547f8 ap_process_request (44b450, 0, 4, 44b450, 0, 443728) + 160 r>=A000051964 ap_process_http_connection (4436c8, 443430, 443430, 0, 998d0,=
2) + 10c

=A000042790 ap_run_process_connection (4436c8, 443430, 443430, 0, 441438, 4=
453f8) + 3c
=A000067a10 child_main (0, 0, 99c00, 99800, 11177, 99800) + =
42c
=A000067c74 make_child (67400, 0, 0, fe781020, 440fc8, 0) + ec
=
=A000067d44 startup_children (5, feebd280, 0, 10, 1cf4, a) + 68

=A0000686b0 ap_mpm_run (a4878, 99800, ab4b8, 99c00, 99800, 99800) + 950
=
=A0000291f4 main     (a4878, 96c00, 99000, 99000, a2870, 0) + 780 r>=A000028594 _start   (0, 0, 0, 0, 0, 0) + 5c

My MPM is straigh=
t pre-fork, default 5 children to start with.=A0 Unfortunately, this stuff =
needs to go into production soon.




On Wed, Jan 27, 2010 at 11:42 AM, To=
sh Cooey <tosh@1=
200group.com
>
wrote:
le=3D"border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex;=
padding-left: 1ex;">
I'm seeing the same thing on my DEV server which is:



Server: Apache/2.2.11 (Ubuntu) mod_perl/2.0.4 Perl/v5.10.0



I see it in the mornings when it has been sitting around all night doing no=
thing, and the first couple hits result in core dumps/seg faults, then afte=
r pounding REFRESH like a mad monkey a couple times it then works.




This is on an Amazon EC2 Ubuntu 9.04 jaunty AMI built by Eric Hammond

=A0=
http://ec2=
ubuntu-group.notlong.com




perl -V:




Summary of my perl5 (revision 5 version 10 subversion 0) configuration:

=A0Platform:

=A0 =A0osname=3Dlinux, osvers=3D2.6.24-23-server, archname=3Di486-linux-gn=
u-thread-multi

=A0 =A0uname=3D'linux rothera 2.6.24-23-server #1 smp wed apr 1 22:22:=
14 utc 2009 i686 gnulinux '

=A0 =A0config_args=3D'-Dusethreads -Duselargefiles -Dccflags=3D-DDEBIA=
N -Dcccdlflags=3D-fPIC -Darchname=3Di486-linux-gnu -Dprefix=3D/usr -Dprivli=
b=3D/usr/share/perl/5.10 -Darchlib=3D/usr/lib/perl/5.10 -Dvendorprefix=3D/u=
sr -Dvendorlib=3D/usr/share/perl5 -Dvendorarch=3D/usr/lib/perl5 -Dsiteprefi=
x=3D/usr/local -Dsitelib=3D/usr/local/share/perl/5.10.0 -Dsitearch=3D/usr/l=
ocal/lib/perl/5.10.0 -Dman1dir=3D/usr/share/man/man1 -Dman3dir=3D/usr/share=
/man/man3 -Dsiteman1dir=3D/usr/local/man/man1 -Dsiteman3dir=3D/usr/local/ma=
n/man3 -Dman1ext=3D1 -Dman3ext=3D3perl -Dpager=3D/usr/bin/sensible-pager -U=
afs -Ud_csh -Ud_ualarm -Uusesfio -Uusenm -DDEBUGGING=3D-g -Doptimize=3D-O2 =
-Duseshrplib -Dlibperl=3Dlibperl.so.5.10.0 -Dd_dosuid -des'
=3D"im">


=A0 =A0hint=3Drecommended, useposix=3Dtrue, d_sigaction=3Ddefine

=A0 =A0useithreads=3Ddefine, usemultiplicity=3Ddefine
>
=A0 =A0useperlio=3Ddefine, d_sfio=3Dundef, uselargefiles=3Ddefine, usesock=
s=3Dundef

=A0 =A0use64bitint=3Dundef, use64bitall=3Dundef, uselongdouble=3Dundef

=A0 =A0usemymalloc=3Dn, bincompat5005=3Dundef

=A0Compiler:

=A0 =A0cc=3D'cc', ccflags =3D'-D_REENTRANT -D_GNU_SOURCE -DDEB=
IAN -fno-strict-aliasing -pipe -I/usr/local/include -D_LARGEFILE_SOURCE -D_=
FILE_OFFSET_BITS=3D64',

=A0 =A0optimize=3D'-O2 -g',

=A0 =A0cppflags=3D'-D_REENTRANT -D_GNU_SOURCE -DDEBIAN -fno-strict-ali=
asing -pipe -I/usr/local/include'

=A0 =A0ccversion=3D'', gccversion=3D'4.3.3', gccosandvers=
=3D''

=A0 =A0intsize=3D4, longsize=3D4, ptrsize=3D4, doublesize=3D8, byteorder=
=3D1234

=A0 =A0d_longlong=3Ddefine, longlongsize=3D8, d_longdbl=3Ddefine, longdbls=
ize=3D12


=A0 =A0ivtype=3D'long', ivsize=3D4, nvtype=3D'double', nvs=
ize=3D8, Off_t=3D'off_t', lseeksize=3D8

=A0 =A0alignbytes=3D4, prototype=3Ddefine

=A0Linker and Libraries:

=A0 =A0ld=3D'cc', ldflags =3D' -L/usr/local/lib'

=A0 =A0libpth=3D/usr/local/lib /lib /usr/lib /usr/lib64

=A0 =A0libs=3D-lgdbm -lgdbm_compat -ldb -ldl -lm -lpthread -lc -lcrypt

=A0 =A0perllibs=3D-ldl -lm -lpthread -lc -lcrypt

=A0 =A0libc=3D/lib/libc-2=
..9.so
, so=3Dso, useshrplib=3Dtrue, libperl=3Dlibperl.so.5.10.0

=A0 =A0gnulibc_version=3D'2.9'

=A0Dynamic Linking:

=A0 =A0dlsrc=3Ddl_dlopen.xs, dlext=3Dso, d_dlsymun=3Dundef, ccdlflags=3D&#=
39;-Wl,-E'

=A0 =A0cccdlflags=3D'-fPIC', lddlflags=3D'-shared -O2 -g -L/us=
r/local/lib'






Characteristics of this binary (from libperl):

=A0Compile-time options: MULTIPLICITY PERL_DONT_CREATE_GVSV

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0PERL_IMPLICIT_CONTEXT PERL_=
MALLOC_WRAP USE_ITHREADS

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0USE_LARGE_FILES USE_PERLIO =
USE_REENTRANT_API

=A0Built under linux

=A0Compiled at Jun 26 2009 18:23:00

=A0@INC:

=A0 =A0/etc/perl

=A0 =A0/usr/local/lib/perl/5.10.0

=A0 =A0/usr/local/share/perl/5.10.0

=A0 =A0/usr/lib/perl5

=A0 =A0/usr/share/perl5

=A0 =A0/usr/lib/perl/5.10

=A0 =A0/usr/share/perl/5.10

=A0 =A0/usr/local/lib/site_perl

=A0 =A0.





Of course this could also just be all the illegitimate bastard apache child=
ren I have been spawning being totally unprepared to answer a real web requ=
est, but I'm keeping an eye on this server...



Tosh






Pas Argenio wrote:

204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Sorry if this is naive, but has anyone gotten this combo to work? =A0Upgrad=
ed to Apache 2.2.14, now crash on almost every request, then next httpd pic=
ks up the request and actually displays the page...



I've got massive core dumps all ending in malloc/free, truss shows last=
operation was close()




--

McIntosh Cooey - Twelve Hundred Group LLC - ..com/" target=3D"_blank">http://www.1200group.com/




--000e0cd5cb3a991f80047e28ba34--

Re: mod_perl-2.0.4 with Apache 2.2.9 and perl 5.10.0 intermittent crashing

am 27.01.2010 19:50:52 von Tosh Cooey

perl -V

I don't think it's an Apache issue, my purely superstitious feelings
based on the ravens I saw flying around this morning is that blame lies
with perl 5.10, I just don't trust even number releases, even less round
numbers cleanly divisible by 10.

But then $r scares me so...

Are you in Europe? It's really cold here, so maybe Apache needs to warm
up first like a car, it takes a couple moments before it turns over?

But in all seriousness I am also having the same problem. How did you
create your setup? From packages or building everything from source?

I'm using packages and suspect fault there as well...

Tosh


Pas Argenio wrote:
> I just fell back to Apache-2.2.8 but same problem. By the way, Tosh,
> how did you dump the libperl info?
> I'm getting more core dumps than you, but pages are displayed by the
> next child to pick it up. Here is the most common stack trace (via
> Solaris' pstack command):
>
> Wed Jan 27 09:06:35 EST 2010
> core '/tmp/core' of 13862: /usr/local/apache/bin/httpd -k start
> fee56ec8 t_delete (528c90, 228, 0, fecb2e24, fef303a8, c20360) + 74
> fee56ac4 realfree (528a60, 229, d995c, fee56eb0, 0, 528a58) + 8c
> fee56610 _malloc_unlocked (528598, 6f8, 528590, ff, 0, 0) + 260
> fee56394 malloc (4c8, 1, da058, fee563a0, fef303a8, fef3a518) + 4c
> fec8ae8c Perl_safesysmalloc (4c8, 0, 2000000, fffffff8, c000004,
> c329d9) + 10
> fecc2294 Perl_sv_grow (d17680, 4c8, 0, 2000000, 4cb, 0) + f8
> fecbd094 Perl_sv_setsv_flags (84408, 4c7, 12c04540, d17680, d177e0,
> ff800000) + ea0
> fecbd5ac Perl_sv_mortalcopy (d177e0, 5e65d0, fed76000, 78, fed74400,
> d17680) + 60
> fecb2e24 Perl_pp_aassign (fed76294, 5b8c20, 5b8c0c, 5b8c1c, 0,
> fed74400) + 154
> fecaa830 Perl_runops_standard (0, fecb2cd0, fed75c88, 91c2c0, fed75c00,
> 91cec0) + 1c
> feca4ce0 Perl_call_sv (fed75c00, fed76000, fed76000, fed75c00,
> fed75c00, 91c2c0) + 4a4
> fec1ca94 modperl_callback (0, 44b410, 44b450, ab4b8, 160080, fed76000)
> + 338
> fec1d05c modperl_callback_run_handlers (fed75dec, 4, 44b450, 121dc8,
> ab4b8, 0) + 31c
> fec1d32c modperl_callback_per_dir (6, 44b450, 1, 1, ffbff03c, 214678) + 24
> fec19ab4 modperl_response_handler_run (44b450, 0, 1460, 0, fed761d8,
> fed76000) + 18
> fec19cc4 modperl_response_handler_cgi (44b450, 4484b8, 121520, 0,
> 99070, 9) + 134
> 0003bfe4 ap_run_handler (44b450, 0, c, 449018, 448728, 0) + 3c
> 0003c460 ap_invoke_handler (44b450, 77400, 44b450, 0, fe780020, 0) + b8
> 000547f8 ap_process_request (44b450, 0, 4, 44b450, 0, 443728) + 160
> 00051964 ap_process_http_connection (4436c8, 443430, 443430, 0, 998d0,
> 2) + 10c
> 00042790 ap_run_process_connection (4436c8, 443430, 443430, 0, 441438,
> 4453f8) + 3c
> 00067a10 child_main (0, 0, 99c00, 99800, 11177, 99800) + 42c
> 00067c74 make_child (67400, 0, 0, fe781020, 440fc8, 0) + ec
> 00067d44 startup_children (5, feebd280, 0, 10, 1cf4, a) + 68
> 000686b0 ap_mpm_run (a4878, 99800, ab4b8, 99c00, 99800, 99800) + 950
> 000291f4 main (a4878, 96c00, 99000, 99000, a2870, 0) + 780
> 00028594 _start (0, 0, 0, 0, 0, 0) + 5c
>
> My MPM is straight pre-fork, default 5 children to start with.
> Unfortunately, this stuff needs to go into production soon.
>
>
>
> On Wed, Jan 27, 2010 at 11:42 AM, Tosh Cooey > > wrote:
>
> I'm seeing the same thing on my DEV server which is:
>
> Server: Apache/2.2.11 (Ubuntu) mod_perl/2.0.4 Perl/v5.10.0
>
> I see it in the mornings when it has been sitting around all night
> doing nothing, and the first couple hits result in core dumps/seg
> faults, then after pounding REFRESH like a mad monkey a couple times
> it then works.
>
> This is on an Amazon EC2 Ubuntu 9.04 jaunty AMI built by Eric Hammond
> http://alestic.com http://ec2ubuntu-group.notlong.com
>
> perl -V:
>
>
> Summary of my perl5 (revision 5 version 10 subversion 0) configuration:
> Platform:
> osname=linux, osvers=2.6.24-23-server,
> archname=i486-linux-gnu-thread-multi
> uname='linux rothera 2.6.24-23-server #1 smp wed apr 1 22:22:14
> utc 2009 i686 gnulinux '
> config_args='-Dusethreads -Duselargefiles -Dccflags=-DDEBIAN
> -Dcccdlflags=-fPIC -Darchname=i486-linux-gnu -Dprefix=/usr
> -Dprivlib=/usr/share/perl/5.10 -Darchlib=/usr/lib/perl/5.10
> -Dvendorprefix=/usr -Dvendorlib=/usr/share/perl5
> -Dvendorarch=/usr/lib/perl5 -Dsiteprefix=/usr/local
> -Dsitelib=/usr/local/share/perl/5.10.0
> -Dsitearch=/usr/local/lib/perl/5.10.0 -Dman1dir=/usr/share/man/man1
> -Dman3dir=/usr/share/man/man3 -Dsiteman1dir=/usr/local/man/man1
> -Dsiteman3dir=/usr/local/man/man3 -Dman1ext=1 -Dman3ext=3perl
> -Dpager=/usr/bin/sensible-pager -Uafs -Ud_csh -Ud_ualarm -Uusesfio
> -Uusenm -DDEBUGGING=-g -Doptimize=-O2 -Duseshrplib
> -Dlibperl=libperl.so.5.10.0 -Dd_dosuid -des'
>
> hint=recommended, useposix=true, d_sigaction=define
> useithreads=define, usemultiplicity=define
>
> useperlio=define, d_sfio=undef, uselargefiles=define, usesocks=undef
> use64bitint=undef, use64bitall=undef, uselongdouble=undef
> usemymalloc=n, bincompat5005=undef
> Compiler:
> cc='cc', ccflags ='-D_REENTRANT -D_GNU_SOURCE -DDEBIAN
> -fno-strict-aliasing -pipe -I/usr/local/include -D_LARGEFILE_SOURCE
> -D_FILE_OFFSET_BITS=64',
> optimize='-O2 -g',
> cppflags='-D_REENTRANT -D_GNU_SOURCE -DDEBIAN
> -fno-strict-aliasing -pipe -I/usr/local/include'
> ccversion='', gccversion='4.3.3', gccosandvers=''
> intsize=4, longsize=4, ptrsize=4, doublesize=8, byteorder=1234
> d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=12
>
> ivtype='long', ivsize=4, nvtype='double', nvsize=8,
> Off_t='off_t', lseeksize=8
> alignbytes=4, prototype=define
> Linker and Libraries:
> ld='cc', ldflags =' -L/usr/local/lib'
> libpth=/usr/local/lib /lib /usr/lib /usr/lib64
> libs=-lgdbm -lgdbm_compat -ldb -ldl -lm -lpthread -lc -lcrypt
> perllibs=-ldl -lm -lpthread -lc -lcrypt
> libc=/lib/libc-2.9.so , so=so,
> useshrplib=true, libperl=libperl.so.5.10.0
> gnulibc_version='2.9'
> Dynamic Linking:
> dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags='-Wl,-E'
> cccdlflags='-fPIC', lddlflags='-shared -O2 -g -L/usr/local/lib'
>
>
>
> Characteristics of this binary (from libperl):
> Compile-time options: MULTIPLICITY PERL_DONT_CREATE_GVSV
> PERL_IMPLICIT_CONTEXT PERL_MALLOC_WRAP
> USE_ITHREADS
> USE_LARGE_FILES USE_PERLIO USE_REENTRANT_API
> Built under linux
> Compiled at Jun 26 2009 18:23:00
> @INC:
> /etc/perl
> /usr/local/lib/perl/5.10.0
> /usr/local/share/perl/5.10.0
> /usr/lib/perl5
> /usr/share/perl5
> /usr/lib/perl/5.10
> /usr/share/perl/5.10
> /usr/local/lib/site_perl
> .
>
>
> Of course this could also just be all the illegitimate bastard
> apache children I have been spawning being totally unprepared to
> answer a real web request, but I'm keeping an eye on this server...
>
> Tosh
>
>
>
> Pas Argenio wrote:
>
> Sorry if this is naive, but has anyone gotten this combo to
> work? Upgraded to Apache 2.2.14, now crash on almost every
> request, then next httpd picks up the request and actually
> displays the page...
>
> I've got massive core dumps all ending in malloc/free, truss
> shows last operation was close()
>
>
> --
> McIntosh Cooey - Twelve Hundred Group LLC - http://www.1200group.com/
>
>

--
McIntosh Cooey - Twelve Hundred Group LLC - http://www.1200group.com/

Re: mod_perl-2.0.4 with Apache 2.2.9 and perl 5.10.0 intermittent

am 27.01.2010 20:05:14 von Pas Argenio

--000e0cd47ed214a17d047e2a15df
Content-Type: text/plain; charset=ISO-8859-1

No, wish I were. East coast US, unseasonably warm (we have more hot air in
general).

I'm building everything from source. I've rebuilt perl a half-dozen times
both static & dynamic. I'm working on building 5.10.1 but it seems so
different in the build defaults.
Still to try: mod_perl.2.0.3 and using perl malloc

On Wed, Jan 27, 2010 at 1:50 PM, Tosh Cooey wrote:

> perl -V
>
> I don't think it's an Apache issue, my purely superstitious feelings based
> on the ravens I saw flying around this morning is that blame lies with perl
> 5.10, I just don't trust even number releases, even less round numbers
> cleanly divisible by 10.
>
> But then $r scares me so...
>
> Are you in Europe? It's really cold here, so maybe Apache needs to warm up
> first like a car, it takes a couple moments before it turns over?
>
> But in all seriousness I am also having the same problem. How did you
> create your setup? From packages or building everything from source?
>
> I'm using packages and suspect fault there as well...
>
> Tosh
>
>
> Pas Argenio wrote:
>
>> I just fell back to Apache-2.2.8 but same problem. By the way, Tosh, how
>> did you dump the libperl info?
>> I'm getting more core dumps than you, but pages are displayed by the next
>> child to pick it up. Here is the most common stack trace (via Solaris'
>> pstack command):
>>
>> Wed Jan 27 09:06:35 EST 2010
>> core '/tmp/core' of 13862: /usr/local/apache/bin/httpd -k start
>> fee56ec8 t_delete (528c90, 228, 0, fecb2e24, fef303a8, c20360) + 74
>> fee56ac4 realfree (528a60, 229, d995c, fee56eb0, 0, 528a58) + 8c
>> fee56610 _malloc_unlocked (528598, 6f8, 528590, ff, 0, 0) + 260
>> fee56394 malloc (4c8, 1, da058, fee563a0, fef303a8, fef3a518) + 4c
>> fec8ae8c Perl_safesysmalloc (4c8, 0, 2000000, fffffff8, c000004, c329d9)
>> + 10
>> fecc2294 Perl_sv_grow (d17680, 4c8, 0, 2000000, 4cb, 0) + f8
>> fecbd094 Perl_sv_setsv_flags (84408, 4c7, 12c04540, d17680, d177e0,
>> ff800000) + ea0
>> fecbd5ac Perl_sv_mortalcopy (d177e0, 5e65d0, fed76000, 78, fed74400,
>> d17680) + 60
>> fecb2e24 Perl_pp_aassign (fed76294, 5b8c20, 5b8c0c, 5b8c1c, 0, fed74400)
>> + 154
>> fecaa830 Perl_runops_standard (0, fecb2cd0, fed75c88, 91c2c0, fed75c00,
>> 91cec0) + 1c
>> feca4ce0 Perl_call_sv (fed75c00, fed76000, fed76000, fed75c00, fed75c00,
>> 91c2c0) + 4a4
>> fec1ca94 modperl_callback (0, 44b410, 44b450, ab4b8, 160080, fed76000) +
>> 338
>> fec1d05c modperl_callback_run_handlers (fed75dec, 4, 44b450, 121dc8,
>> ab4b8, 0) + 31c
>> fec1d32c modperl_callback_per_dir (6, 44b450, 1, 1, ffbff03c, 214678) +
>> 24
>> fec19ab4 modperl_response_handler_run (44b450, 0, 1460, 0, fed761d8,
>> fed76000) + 18
>> fec19cc4 modperl_response_handler_cgi (44b450, 4484b8, 121520, 0, 99070,
>> 9) + 134
>> 0003bfe4 ap_run_handler (44b450, 0, c, 449018, 448728, 0) + 3c
>> 0003c460 ap_invoke_handler (44b450, 77400, 44b450, 0, fe780020, 0) + b8
>> 000547f8 ap_process_request (44b450, 0, 4, 44b450, 0, 443728) + 160
>> 00051964 ap_process_http_connection (4436c8, 443430, 443430, 0, 998d0, 2)
>> + 10c
>> 00042790 ap_run_process_connection (4436c8, 443430, 443430, 0, 441438,
>> 4453f8) + 3c
>> 00067a10 child_main (0, 0, 99c00, 99800, 11177, 99800) + 42c
>> 00067c74 make_child (67400, 0, 0, fe781020, 440fc8, 0) + ec
>> 00067d44 startup_children (5, feebd280, 0, 10, 1cf4, a) + 68
>> 000686b0 ap_mpm_run (a4878, 99800, ab4b8, 99c00, 99800, 99800) + 950
>> 000291f4 main (a4878, 96c00, 99000, 99000, a2870, 0) + 780
>> 00028594 _start (0, 0, 0, 0, 0, 0) + 5c
>>
>> My MPM is straight pre-fork, default 5 children to start with.
>> Unfortunately, this stuff needs to go into production soon.
>>
>>
>>
>> On Wed, Jan 27, 2010 at 11:42 AM, Tosh Cooey >> tosh@1200group.com>> wrote:
>>
>> I'm seeing the same thing on my DEV server which is:
>>
>> Server: Apache/2.2.11 (Ubuntu) mod_perl/2.0.4 Perl/v5.10.0
>>
>> I see it in the mornings when it has been sitting around all night
>> doing nothing, and the first couple hits result in core dumps/seg
>> faults, then after pounding REFRESH like a mad monkey a couple times
>> it then works.
>>
>> This is on an Amazon EC2 Ubuntu 9.04 jaunty AMI built by Eric Hammond
>> http://alestic.com http://ec2ubuntu-group.notlong.com
>>
>> perl -V:
>>
>>
>> Summary of my perl5 (revision 5 version 10 subversion 0) configuration:
>> Platform:
>> osname=linux, osvers=2.6.24-23-server,
>> archname=i486-linux-gnu-thread-multi
>> uname='linux rothera 2.6.24-23-server #1 smp wed apr 1 22:22:14
>> utc 2009 i686 gnulinux '
>> config_args='-Dusethreads -Duselargefiles -Dccflags=-DDEBIAN
>> -Dcccdlflags=-fPIC -Darchname=i486-linux-gnu -Dprefix=/usr
>> -Dprivlib=/usr/share/perl/5.10 -Darchlib=/usr/lib/perl/5.10
>> -Dvendorprefix=/usr -Dvendorlib=/usr/share/perl5
>> -Dvendorarch=/usr/lib/perl5 -Dsiteprefix=/usr/local
>> -Dsitelib=/usr/local/share/perl/5.10.0
>> -Dsitearch=/usr/local/lib/perl/5.10.0 -Dman1dir=/usr/share/man/man1
>> -Dman3dir=/usr/share/man/man3 -Dsiteman1dir=/usr/local/man/man1
>> -Dsiteman3dir=/usr/local/man/man3 -Dman1ext=1 -Dman3ext=3perl
>> -Dpager=/usr/bin/sensible-pager -Uafs -Ud_csh -Ud_ualarm -Uusesfio
>> -Uusenm -DDEBUGGING=-g -Doptimize=-O2 -Duseshrplib
>> -Dlibperl=libperl.so.5.10.0 -Dd_dosuid -des'
>>
>> hint=recommended, useposix=true, d_sigaction=define
>> useithreads=define, usemultiplicity=define
>>
>> useperlio=define, d_sfio=undef, uselargefiles=define, usesocks=undef
>> use64bitint=undef, use64bitall=undef, uselongdouble=undef
>> usemymalloc=n, bincompat5005=undef
>> Compiler:
>> cc='cc', ccflags ='-D_REENTRANT -D_GNU_SOURCE -DDEBIAN
>> -fno-strict-aliasing -pipe -I/usr/local/include -D_LARGEFILE_SOURCE
>> -D_FILE_OFFSET_BITS=64',
>> optimize='-O2 -g',
>> cppflags='-D_REENTRANT -D_GNU_SOURCE -DDEBIAN
>> -fno-strict-aliasing -pipe -I/usr/local/include'
>> ccversion='', gccversion='4.3.3', gccosandvers=''
>> intsize=4, longsize=4, ptrsize=4, doublesize=8, byteorder=1234
>> d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=12
>>
>> ivtype='long', ivsize=4, nvtype='double', nvsize=8,
>> Off_t='off_t', lseeksize=8
>> alignbytes=4, prototype=define
>> Linker and Libraries:
>> ld='cc', ldflags =' -L/usr/local/lib'
>> libpth=/usr/local/lib /lib /usr/lib /usr/lib64
>> libs=-lgdbm -lgdbm_compat -ldb -ldl -lm -lpthread -lc -lcrypt
>> perllibs=-ldl -lm -lpthread -lc -lcrypt
>> libc=/lib/libc-2.9.so , so=so,
>>
>> useshrplib=true, libperl=libperl.so.5.10.0
>> gnulibc_version='2.9'
>> Dynamic Linking:
>> dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags='-Wl,-E'
>> cccdlflags='-fPIC', lddlflags='-shared -O2 -g -L/usr/local/lib'
>>
>>
>>
>> Characteristics of this binary (from libperl):
>> Compile-time options: MULTIPLICITY PERL_DONT_CREATE_GVSV
>> PERL_IMPLICIT_CONTEXT PERL_MALLOC_WRAP
>> USE_ITHREADS
>> USE_LARGE_FILES USE_PERLIO USE_REENTRANT_API
>> Built under linux
>> Compiled at Jun 26 2009 18:23:00
>> @INC:
>> /etc/perl
>> /usr/local/lib/perl/5.10.0
>> /usr/local/share/perl/5.10.0
>> /usr/lib/perl5
>> /usr/share/perl5
>> /usr/lib/perl/5.10
>> /usr/share/perl/5.10
>> /usr/local/lib/site_perl
>> .
>>
>>
>> Of course this could also just be all the illegitimate bastard
>> apache children I have been spawning being totally unprepared to
>> answer a real web request, but I'm keeping an eye on this server...
>>
>> Tosh
>>
>>
>>
>> Pas Argenio wrote:
>>
>> Sorry if this is naive, but has anyone gotten this combo to
>> work? Upgraded to Apache 2.2.14, now crash on almost every
>> request, then next httpd picks up the request and actually
>> displays the page...
>>
>> I've got massive core dumps all ending in malloc/free, truss
>> shows last operation was close()
>>
>>
>> -- McIntosh Cooey - Twelve Hundred Group LLC -
>> http://www.1200group.com/
>>
>>
>>
> --
> McIntosh Cooey - Twelve Hundred Group LLC - http://www.1200group.com/
>

--000e0cd47ed214a17d047e2a15df
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

No, wish I were. East coast US, unseasonably warm (we have more hot air in =
general).

I'm building everything from source.=A0 I've rebui=
lt perl a half-dozen times both static & dynamic.=A0 I'm working on=
building 5.10.1 but it seems so different in the build defaults.

Still to try: mod_perl.2.0.3 and using perl malloc

il_quote">On Wed, Jan 27, 2010 at 1:50 PM, Tosh Cooey <=
;>
wr=
ote:

204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">perl -V



I don't think it's an Apache issue, my purely superstitious feeling=
s based on the ravens I saw flying around this morning is that blame lies w=
ith perl 5.10, I just don't trust even number releases, even less round=
numbers cleanly divisible by 10.




But then $r scares me so...



Are you in Europe? =A0It's really cold here, so maybe Apache needs to w=
arm up first like a car, it takes a couple moments before it turns over? >


But in all seriousness I am also having the same problem. =A0How did you cr=
eate your setup? =A0From packages or building everything from source?



I'm using packages and suspect fault there as well...



Tosh





Pas Argenio wrote:

204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<=
div class=3D"h5">
I just fell back to Apache-2.2.8 but same problem. =A0By the way, Tosh, how=
did you dump the libperl info?

I'm getting more core dumps than you, but pages are displayed by the ne=
xt child to pick it up. =A0Here is the most common stack trace (via Solaris=
' pstack command):



Wed Jan 27 09:06:35 EST 2010

core '/tmp/core' of 13862: =A0 =A0/usr/local/apache/bin/httpd -k st=
art

=A0fee56ec8 t_delete (528c90, 228, 0, fecb2e24, fef303a8, c20360) + 74

=A0fee56ac4 realfree (528a60, 229, d995c, fee56eb0, 0, 528a58) + 8c

=A0fee56610 _malloc_unlocked (528598, 6f8, 528590, ff, 0, 0) + 260

=A0fee56394 malloc =A0 (4c8, 1, da058, fee563a0, fef303a8, fef3a518) + 4c r>
=A0fec8ae8c Perl_safesysmalloc (4c8, 0, 2000000, fffffff8, c000004, c329d9)=
+ 10

=A0fecc2294 Perl_sv_grow (d17680, 4c8, 0, 2000000, 4cb, 0) + f8

=A0fecbd094 Perl_sv_setsv_flags (84408, 4c7, 12c04540, d17680, d177e0, ff80=
0000) + ea0

=A0fecbd5ac Perl_sv_mortalcopy (d177e0, 5e65d0, fed76000, 78, fed74400, d17=
680) + 60

=A0fecb2e24 Perl_pp_aassign (fed76294, 5b8c20, 5b8c0c, 5b8c1c, 0, fed74400)=
+ 154

=A0fecaa830 Perl_runops_standard (0, fecb2cd0, fed75c88, 91c2c0, fed75c00, =
91cec0) + 1c

=A0feca4ce0 Perl_call_sv (fed75c00, fed76000, fed76000, fed75c00, fed75c00,=
91c2c0) + 4a4

=A0fec1ca94 modperl_callback (0, 44b410, 44b450, ab4b8, 160080, fed76000) +=
338

=A0fec1d05c modperl_callback_run_handlers (fed75dec, 4, 44b450, 121dc8, ab4=
b8, 0) + 31c

=A0fec1d32c modperl_callback_per_dir (6, 44b450, 1, 1, ffbff03c, 214678) + =
24

=A0fec19ab4 modperl_response_handler_run (44b450, 0, 1460, 0, fed761d8, fed=
76000) + 18

=A0fec19cc4 modperl_response_handler_cgi (44b450, 4484b8, 121520, 0, 99070,=
9) + 134

=A00003bfe4 ap_run_handler (44b450, 0, c, 449018, 448728, 0) + 3c

=A00003c460 ap_invoke_handler (44b450, 77400, 44b450, 0, fe780020, 0) + b8<=
br>
=A0000547f8 ap_process_request (44b450, 0, 4, 44b450, 0, 443728) + 160

=A000051964 ap_process_http_connection (4436c8, 443430, 443430, 0, 998d0, 2=
) + 10c

=A000042790 ap_run_process_connection (4436c8, 443430, 443430, 0, 441438, 4=
453f8) + 3c

=A000067a10 child_main (0, 0, 99c00, 99800, 11177, 99800) + 42c

=A000067c74 make_child (67400, 0, 0, fe781020, 440fc8, 0) + ec

=A000067d44 startup_children (5, feebd280, 0, 10, 1cf4, a) + 68

=A0000686b0 ap_mpm_run (a4878, 99800, ab4b8, 99c00, 99800, 99800) + 950

=A0000291f4 main =A0 =A0 (a4878, 96c00, 99000, 99000, a2870, 0) + 780

=A000028594 _start =A0 (0, 0, 0, 0, 0, 0) + 5c



My MPM is straight pre-fork, default 5 children to start with. =A0Unfortuna=
tely, this stuff needs to go into production soon.







On Wed, Jan 27, 2010 at 11:42 AM, Tosh Cooey < 0group.com" target=3D"_blank">tosh@1200group.com <mailto: mailto:tosh@1200group.com" target=3D"_blank">tosh@1200group.com>>=
wrote:




=A0 =A0I'm seeing the same thing on my DEV server which is:



=A0 =A0Server: Apache/2.2.11 (Ubuntu) mod_perl/2.0.4 Perl/v5.10.0



=A0 =A0I see it in the mornings when it has been sitting around all night<=
br>
=A0 =A0doing nothing, and the first couple hits result in core dumps/seg r>
=A0 =A0faults, then after pounding REFRESH like a mad monkey a couple time=
s

=A0 =A0it then works.



=A0 =A0This is on an Amazon EC2 Ubuntu 9.04 jaunty AMI built by Eric Hammo=
nd

=A0 =A0http://alestic.com=
=A0ht=
tp://ec2ubuntu-group.notlong.com




=A0 =A0perl -V:





=A0 =A0Summary of my perl5 (revision 5 version 10 subversion 0) configurat=
ion:

=A0 =A0 Platform:

=A0 =A0 =A0 osname=3Dlinux, osvers=3D2.6.24-23-server,

=A0 =A0archname=3Di486-linux-gnu-thread-multi

=A0 =A0 =A0 uname=3D'linux rothera 2.6.24-23-server #1 smp wed apr 1 2=
2:22:14

=A0 =A0utc 2009 i686 gnulinux '

=A0 =A0 =A0 config_args=3D'-Dusethreads -Duselargefiles -Dccflags=3D-D=
DEBIAN

=A0 =A0-Dcccdlflags=3D-fPIC -Darchname=3Di486-linux-gnu -Dprefix=3D/usr >
=A0 =A0-Dprivlib=3D/usr/share/perl/5.10 -Darchlib=3D/usr/lib/perl/5.10

=A0 =A0-Dvendorprefix=3D/usr -Dvendorlib=3D/usr/share/perl5

=A0 =A0-Dvendorarch=3D/usr/lib/perl5 -Dsiteprefix=3D/usr/local

=A0 =A0-Dsitelib=3D/usr/local/share/perl/5.10.0

=A0 =A0-Dsitearch=3D/usr/local/lib/perl/5.10.0 -Dman1dir=3D/usr/share/man/=
man1

=A0 =A0-Dman3dir=3D/usr/share/man/man3 -Dsiteman1dir=3D/usr/local/man/man1=


=A0 =A0-Dsiteman3dir=3D/usr/local/man/man3 -Dman1ext=3D1 -Dman3ext=3D3perl=


=A0 =A0-Dpager=3D/usr/bin/sensible-pager -Uafs -Ud_csh -Ud_ualarm -Uusesfi=
o

=A0 =A0-Uusenm -DDEBUGGING=3D-g -Doptimize=3D-O2 -Duseshrplib

=A0 =A0-Dlibperl=3Dlibperl.so.5.10.0 -Dd_dosuid -des'



=A0 =A0 =A0 hint=3Drecommended, useposix=3Dtrue, d_sigaction=3Ddefine

=A0 =A0 =A0 useithreads=3Ddefine, usemultiplicity=3Ddefine



=A0 =A0 =A0 useperlio=3Ddefine, d_sfio=3Dundef, uselargefiles=3Ddefine, us=
esocks=3Dundef

=A0 =A0 =A0 use64bitint=3Dundef, use64bitall=3Dundef, uselongdouble=3Dunde=
f

=A0 =A0 =A0 usemymalloc=3Dn, bincompat5005=3Dundef

=A0 =A0 Compiler:

=A0 =A0 =A0 cc=3D'cc', ccflags =3D'-D_REENTRANT -D_GNU_SOURCE =
-DDEBIAN

=A0 =A0-fno-strict-aliasing -pipe -I/usr/local/include -D_LARGEFILE_SOURCE=


=A0 =A0-D_FILE_OFFSET_BITS=3D64',

=A0 =A0 =A0 optimize=3D'-O2 -g',

=A0 =A0 =A0 cppflags=3D'-D_REENTRANT -D_GNU_SOURCE -DDEBIAN

=A0 =A0-fno-strict-aliasing -pipe -I/usr/local/include'

=A0 =A0 =A0 ccversion=3D'', gccversion=3D'4.3.3', gccosand=
vers=3D''

=A0 =A0 =A0 intsize=3D4, longsize=3D4, ptrsize=3D4, doublesize=3D8, byteor=
der=3D1234

=A0 =A0 =A0 d_longlong=3Ddefine, longlongsize=3D8, d_longdbl=3Ddefine, lon=
gdblsize=3D12



=A0 =A0 =A0 ivtype=3D'long', ivsize=3D4, nvtype=3D'double'=
, nvsize=3D8,

=A0 =A0Off_t=3D'off_t', lseeksize=3D8

=A0 =A0 =A0 alignbytes=3D4, prototype=3Ddefine

=A0 =A0 Linker and Libraries:

=A0 =A0 =A0 ld=3D'cc', ldflags =3D' -L/usr/local/lib'

=A0 =A0 =A0 libpth=3D/usr/local/lib /lib /usr/lib /usr/lib64

=A0 =A0 =A0 libs=3D-lgdbm -lgdbm_compat -ldb -ldl -lm -lpthread -lc -lcryp=
t

=A0 =A0 =A0 perllibs=3D-ldl -lm -lpthread -lc -lcrypt

=A0 =A0 =A0 libc=3D/lib/l=
ibc-2.9.so
<http://=
libc-2.9.so
>, so=3Dso,


=A0 =A0useshrplib=3Dtrue, libperl=3Dlibperl.so.5.10.0

=A0 =A0 =A0 gnulibc_version=3D'2.9'

=A0 =A0 Dynamic Linking:

=A0 =A0 =A0 dlsrc=3Ddl_dlopen.xs, dlext=3Dso, d_dlsymun=3Dundef, ccdlflags=
=3D'-Wl,-E'

=A0 =A0 =A0 cccdlflags=3D'-fPIC', lddlflags=3D'-shared -O2 -g =
-L/usr/local/lib'







=A0 =A0Characteristics of this binary (from libperl):

=A0 =A0 Compile-time options: MULTIPLICITY PERL_DONT_CREATE_GVSV

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 PERL_IMPLICIT_CONTEXT =
PERL_MALLOC_WRAP

=A0 =A0USE_ITHREADS

=A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 USE_LARGE_FILES USE_PE=
RLIO USE_REENTRANT_API

=A0 =A0 Built under linux

=A0 =A0 Compiled at Jun 26 2009 18:23:00

=A0 =A0 @INC:

=A0 =A0 =A0 /etc/perl

=A0 =A0 =A0 /usr/local/lib/perl/5.10.0

=A0 =A0 =A0 /usr/local/share/perl/5.10.0

=A0 =A0 =A0 /usr/lib/perl5

=A0 =A0 =A0 /usr/share/perl5

=A0 =A0 =A0 /usr/lib/perl/5.10

=A0 =A0 =A0 /usr/share/perl/5.10

=A0 =A0 =A0 /usr/local/lib/site_perl

=A0 =A0 =A0 .





=A0 =A0Of course this could also just be all the illegitimate bastard

=A0 =A0apache children I have been spawning being totally unprepared to >
=A0 =A0answer a real web request, but I'm keeping an eye on this serve=
r...



=A0 =A0Tosh







=A0 =A0Pas Argenio wrote:



=A0 =A0 =A0 =A0Sorry if this is naive, but has anyone gotten this combo to=


=A0 =A0 =A0 =A0work? =A0Upgraded to Apache 2.2.14, now crash on almost eve=
ry

=A0 =A0 =A0 =A0request, then next httpd picks up the request and actually<=
br>
=A0 =A0 =A0 =A0displays the page...



=A0 =A0 =A0 =A0I've got massive core dumps all ending in malloc/free, =
truss

=A0 =A0 =A0 =A0shows last operation was close()





=A0 =A0-- =A0 =A0McIntosh Cooey - Twelve Hundred Group LLC - ttp://www.1200group.com/" target=3D"_blank">http://www.1200group.com/ r>







--

McIntosh Cooey - Twelve Hundred Group LLC - ..com/" target=3D"_blank">http://www.1200group.com/




--000e0cd47ed214a17d047e2a15df--

Re: mod_perl-2.0.4 with Apache 2.2.9 and perl 5.10.0 intermittent

am 27.01.2010 20:12:41 von Fred Moyer

Perl 5.10.1 is still relatively new, so you might want to give 5.8.9 a
shot. That is more heavily vetted on lesser used platforms such as
Solaris.

On Wed, Jan 27, 2010 at 9:28 AM, Pas Argenio wrote:
> I just fell back to Apache-2.2.8 but same problem.=A0 By the way, Tosh, h=
ow
> did you dump the libperl info?
> I'm getting more core dumps than you, but pages are displayed by the next
> child to pick it up.=A0 Here is the most common stack trace (via Solaris'
> pstack command):
>
> Wed Jan 27 09:06:35 EST 2010
> core '/tmp/core' of 13862:  =A0 /usr/local/apache/bin/httpd -k start
> =A0fee56ec8 t_delete (528c90, 228, 0, fecb2e24, fef303a8, c20360) + 74
> =A0fee56ac4 realfree (528a60, 229, d995c, fee56eb0, 0, 528a58) + 8c
> =A0fee56610 _malloc_unlocked (528598, 6f8, 528590, ff, 0, 0) + 260
> =A0fee56394 malloc   (4c8, 1, da058, fee563a0, fef303a8, fef3a518) + =
4c
> =A0fec8ae8c Perl_safesysmalloc (4c8, 0, 2000000, fffffff8, c000004, c329d=
9) +
> 10
> =A0fecc2294 Perl_sv_grow (d17680, 4c8, 0, 2000000, 4cb, 0) + f8
> =A0fecbd094 Perl_sv_setsv_flags (84408, 4c7, 12c04540, d17680, d177e0,
> ff800000) + ea0
> =A0fecbd5ac Perl_sv_mortalcopy (d177e0, 5e65d0, fed76000, 78, fed74400,
> d17680) + 60
> =A0fecb2e24 Perl_pp_aassign (fed76294, 5b8c20, 5b8c0c, 5b8c1c, 0, fed7440=
0) +
> 154
> =A0fecaa830 Perl_runops_standard (0, fecb2cd0, fed75c88, 91c2c0, fed75c00=
,
> 91cec0) + 1c
> =A0feca4ce0 Perl_call_sv (fed75c00, fed76000, fed76000, fed75c00, fed75c0=
0,
> 91c2c0) + 4a4
> =A0fec1ca94 modperl_callback (0, 44b410, 44b450, ab4b8, 160080, fed76000)=
+
> 338
> =A0fec1d05c modperl_callback_run_handlers (fed75dec, 4, 44b450, 121dc8, a=
b4b8,
> 0) + 31c
> =A0fec1d32c modperl_callback_per_dir (6, 44b450, 1, 1, ffbff03c, 214678) =
+ 24
> =A0fec19ab4 modperl_response_handler_run (44b450, 0, 1460, 0, fed761d8,
> fed76000) + 18
> =A0fec19cc4 modperl_response_handler_cgi (44b450, 4484b8, 121520, 0, 9907=
0, 9)
> + 134
> =A00003bfe4 ap_run_handler (44b450, 0, c, 449018, 448728, 0) + 3c
> =A00003c460 ap_invoke_handler (44b450, 77400, 44b450, 0, fe780020, 0) + b=
8
> =A0000547f8 ap_process_request (44b450, 0, 4, 44b450, 0, 443728) + 160
> =A000051964 ap_process_http_connection (4436c8, 443430, 443430, 0, 998d0,=
2) +
> 10c
> =A000042790 ap_run_process_connection (4436c8, 443430, 443430, 0, 441438,
> 4453f8) + 3c
> =A000067a10 child_main (0, 0, 99c00, 99800, 11177, 99800) + 42c
> =A000067c74 make_child (67400, 0, 0, fe781020, 440fc8, 0) + ec
> =A000067d44 startup_children (5, feebd280, 0, 10, 1cf4, a) + 68
> =A0000686b0 ap_mpm_run (a4878, 99800, ab4b8, 99c00, 99800, 99800) + 950
> =A0000291f4 main     (a4878, 96c00, 99000, 99000, a2870, 0) + 780
> =A000028594 _start   (0, 0, 0, 0, 0, 0) + 5c
>
> My MPM is straight pre-fork, default 5 children to start with.
> Unfortunately, this stuff needs to go into production soon.
>
>
>
> On Wed, Jan 27, 2010 at 11:42 AM, Tosh Cooey wrote:
>>
>> I'm seeing the same thing on my DEV server which is:
>>
>> Server: Apache/2.2.11 (Ubuntu) mod_perl/2.0.4 Perl/v5.10.0
>>
>> I see it in the mornings when it has been sitting around all night doing
>> nothing, and the first couple hits result in core dumps/seg faults, then
>> after pounding REFRESH like a mad monkey a couple times it then works.
>>
>> This is on an Amazon EC2 Ubuntu 9.04 jaunty AMI built by Eric Hammond
>> http://alestic.com =A0http://ec2ubuntu-group.notlong.com
>>
>> perl -V:
>>
>> Summary of my perl5 (revision 5 version 10 subversion 0) configuration:
>> =A0Platform:
>> =A0 =A0osname=3Dlinux, osvers=3D2.6.24-23-server,
>> archname=3Di486-linux-gnu-thread-multi
>> =A0 =A0uname=3D'linux rothera 2.6.24-23-server #1 smp wed apr 1 22:22:14=
utc
>> 2009 i686 gnulinux '
>> =A0 =A0config_args=3D'-Dusethreads -Duselargefiles -Dccflags=3D-DDEBIAN
>> -Dcccdlflags=3D-fPIC -Darchname=3Di486-linux-gnu -Dprefix=3D/usr
>> -Dprivlib=3D/usr/share/perl/5.10 -Darchlib=3D/usr/lib/perl/5.10
>> -Dvendorprefix=3D/usr -Dvendorlib=3D/usr/share/perl5 -Dvendorarch=3D/usr=
/lib/perl5
>> -Dsiteprefix=3D/usr/local -Dsitelib=3D/usr/local/share/perl/5.10.0
>> -Dsitearch=3D/usr/local/lib/perl/5.10.0 -Dman1dir=3D/usr/share/man/man1
>> -Dman3dir=3D/usr/share/man/man3 -Dsiteman1dir=3D/usr/local/man/man1
>> -Dsiteman3dir=3D/usr/local/man/man3 -Dman1ext=3D1 -Dman3ext=3D3perl
>> -Dpager=3D/usr/bin/sensible-pager -Uafs -Ud_csh -Ud_ualarm -Uusesfio -Uu=
senm
>> -DDEBUGGING=3D-g -Doptimize=3D-O2 -Duseshrplib -Dlibperl=3Dlibperl.so.5.=
10.0
>> -Dd_dosuid -des'
>> =A0 =A0hint=3Drecommended, useposix=3Dtrue, d_sigaction=3Ddefine
>> =A0 =A0useithreads=3Ddefine, usemultiplicity=3Ddefine
>> =A0 =A0useperlio=3Ddefine, d_sfio=3Dundef, uselargefiles=3Ddefine, useso=
cks=3Dundef
>> =A0 =A0use64bitint=3Dundef, use64bitall=3Dundef, uselongdouble=3Dundef
>> =A0 =A0usemymalloc=3Dn, bincompat5005=3Dundef
>> =A0Compiler:
>> =A0 =A0cc=3D'cc', ccflags =3D'-D_REENTRANT -D_GNU_SOURCE -DDEBIAN
>> -fno-strict-aliasing -pipe -I/usr/local/include -D_LARGEFILE_SOURCE
>> -D_FILE_OFFSET_BITS=3D64',
>> =A0 =A0optimize=3D'-O2 -g',
>> =A0 =A0cppflags=3D'-D_REENTRANT -D_GNU_SOURCE -DDEBIAN -fno-strict-alias=
ing
>> -pipe -I/usr/local/include'
>> =A0 =A0ccversion=3D'', gccversion=3D'4.3.3', gccosandvers=3D''
>> =A0 =A0intsize=3D4, longsize=3D4, ptrsize=3D4, doublesize=3D8, byteorder=
=3D1234
>> =A0 =A0d_longlong=3Ddefine, longlongsize=3D8, d_longdbl=3Ddefine, longdb=
lsize=3D12
>> =A0 =A0ivtype=3D'long', ivsize=3D4, nvtype=3D'double', nvsize=3D8, Off_t=
=3D'off_t',
>> lseeksize=3D8
>> =A0 =A0alignbytes=3D4, prototype=3Ddefine
>> =A0Linker and Libraries:
>> =A0 =A0ld=3D'cc', ldflags =3D' -L/usr/local/lib'
>> =A0 =A0libpth=3D/usr/local/lib /lib /usr/lib /usr/lib64
>> =A0 =A0libs=3D-lgdbm -lgdbm_compat -ldb -ldl -lm -lpthread -lc -lcrypt
>> =A0 =A0perllibs=3D-ldl -lm -lpthread -lc -lcrypt
>> =A0 =A0libc=3D/lib/libc-2.9.so, so=3Dso, useshrplib=3Dtrue,
>> libperl=3Dlibperl.so.5.10.0
>> =A0 =A0gnulibc_version=3D'2.9'
>> =A0Dynamic Linking:
>> =A0 =A0dlsrc=3Ddl_dlopen.xs, dlext=3Dso, d_dlsymun=3Dundef, ccdlflags=3D=
'-Wl,-E'
>> =A0 =A0cccdlflags=3D'-fPIC', lddlflags=3D'-shared -O2 -g -L/usr/local/li=
b'
>>
>>
>> Characteristics of this binary (from libperl):
>> =A0Compile-time options: MULTIPLICITY PERL_DONT_CREATE_GVSV
>> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0PERL_IMPLICIT_CONTEXT PER=
L_MALLOC_WRAP USE_ITHREADS
>> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0USE_LARGE_FILES USE_PERLI=
O USE_REENTRANT_API
>> =A0Built under linux
>> =A0Compiled at Jun 26 2009 18:23:00
>> =A0@INC:
>> =A0 =A0/etc/perl
>> =A0 =A0/usr/local/lib/perl/5.10.0
>> =A0 =A0/usr/local/share/perl/5.10.0
>> =A0 =A0/usr/lib/perl5
>> =A0 =A0/usr/share/perl5
>> =A0 =A0/usr/lib/perl/5.10
>> =A0 =A0/usr/share/perl/5.10
>> =A0 =A0/usr/local/lib/site_perl
>> =A0 =A0.
>>
>>
>> Of course this could also just be all the illegitimate bastard apache
>> children I have been spawning being totally unprepared to answer a real =
web
>> request, but I'm keeping an eye on this server...
>>
>> Tosh
>>
>>
>> Pas Argenio wrote:
>>>
>>> Sorry if this is naive, but has anyone gotten this combo to work?
>>> =A0Upgraded to Apache 2.2.14, now crash on almost every request, then n=
ext
>>> httpd picks up the request and actually displays the page...
>>>
>>> I've got massive core dumps all ending in malloc/free, truss shows last
>>> operation was close()
>>
>> --
>> McIntosh Cooey - Twelve Hundred Group LLC - http://www.1200group.com/
>
>

Re: mod_perl-2.0.4 with Apache 2.2.9 and perl 5.10.0 intermittent crashing

am 28.01.2010 02:58:29 von John D Groenveld

In message <30a2344a1001271105t6c690071qb72eb53df2a68456@mail.gmail.com>, Pas A
rgenio writes:
>I'm building everything from source. I've rebuilt perl a half-dozen times
>both static & dynamic. I'm working on building 5.10.1 but it seems so

Does modperl's make test succeed?

Under Solaris, you need to be very careful to match your Perl build
to your Apache.

These are my notes for 64-bit Apache stack for Solaris 10:

Some modperl and libapreq2 tests are failing, but not core dumping.

John
groenveld@acm.org

Re: mod_perl-2.0.4 with Apache 2.2.9 and perl 5.10.0 intermittent

am 28.01.2010 17:31:37 von Pas Argenio

--000e0cd6308a9a9c6e047e3c0d76
Content-Type: text/plain; charset=ISO-8859-1

Thanks John. Currently, I'm building 32-bit for a Sparc platform and also
with man CPAN modules such as DBI and DBD.

I just rebuilt everything carefully, but still coring often before then
displaying a page:
Apache 2.2.8 Perl 5.10.0 mod_perl 2.0.3

I also turned on perl internal malloc

Should I try 64-bit versions next, or fall back to perl 5.8.9?


On Wed, Jan 27, 2010 at 8:58 PM, John D Groenveld
wrote:

> In message <30a2344a1001271105t6c690071qb72eb53df2a68456@mail.gmail.com>,
> Pas A
> rgenio writes:
> >I'm building everything from source. I've rebuilt perl a half-dozen times
> >both static & dynamic. I'm working on building 5.10.1 but it seems so
>
> Does modperl's make test succeed?
>
> Under Solaris, you need to be very careful to match your Perl build
> to your Apache.
>
> These are my notes for 64-bit Apache stack for Solaris 10:
>
> >
> Some modperl and libapreq2 tests are failing, but not core dumping.
>
> John
> groenveld@acm.org
>
>

--000e0cd6308a9a9c6e047e3c0d76
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

Thanks John.=A0 Currently, I'm building 32-bit for a Sparc platform and=
also with man CPAN modules such as DBI and DBD.

I just rebuilt ever=
ything carefully, but still coring often before then displaying a page:

   Apache 2.2.8   Perl 5.10.0   mod_perl 2.0.3

I also tu=
rned on perl internal malloc

Should I try 64-bit versions next, or f=
all back to perl 5.8.9?


On Wed, Jan 2=
7, 2010 at 8:58 PM, John D Groenveld < o:jdg117@elvis.arl.psu.edu">jdg117@elvis.arl.psu.edu> wrote:<=
br>
204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">In message < href=3D"mailto:30a2344a1001271105t6c690071qb72eb53df2a68456@ mail.gmail.com=
">30a2344a1001271105t6c690071qb72eb53df2a68456@mail.gmail.co m>, Pas =
A


rgenio writes:

>I'm building everything from source. =A0I've rebuilt perl a hal=
f-dozen times

>both static & dynamic. =A0I'm working on building 5.10.1 but it=
seems so



Does modperl's make test succeed?



Under Solaris, you need to be very careful to match your Perl build

to your Apache.



These are my notes for 64-bit Apache stack for Solaris 10:

<URL: =3D"_blank">http://www.cse.psu.edu/~groenvel/apache2.html>

Some modperl and libapreq2 tests are failing, but not core dumping.



John








--000e0cd6308a9a9c6e047e3c0d76--

Re: mod_perl-2.0.4 with Apache 2.2.9 and perl 5.10.0 intermittent

am 28.01.2010 17:44:49 von demerphq

2010/1/28 Pas Argenio :
> Thanks John.=A0 Currently, I'm building 32-bit for a Sparc platform and a=
lso
> with man CPAN modules such as DBI and DBD.
>
> I just rebuilt everything carefully, but still coring often before then
> displaying a page:
>    Apache 2.2.8   Perl 5.10.0   mod_perl 2.0.3
>
> I also turned on perl internal malloc
>
> Should I try 64-bit versions next, or fall back to perl 5.8.9?

I recommend you try 5.8.9

Yves





--=20
perl -Mre=3Ddebug -e "/just|another|perl|hacker/"

Re: mod_perl-2.0.4 with Apache 2.2.9 and perl 5.10.0 intermittent

am 28.01.2010 22:30:04 von Pas Argenio

--000e0cd70d0eeb7b4d047e403891
Content-Type: text/plain; charset=ISO-8859-1

I resolved it, finally. Stupidly, we had created a symbolic link:
libmysqlclient.so.15 -> libmysqlclient.so.16.0.0

(I guess our source build of mysql-5.1.32 only provides
libmysqlclient.so.16.0.0 but DBD-mysql-4.008 needs libmysqlclient.so.15)
The fix was to get the real libmysqlclient.so.15 library from the
mysql.5.0.67 Sun package. We should have known software is rarely FORWARD
compatible!
Now to fix it the right way, maybe the latest DBD-mysql uses the .16
library.

Anyway perl 5.10.0 does work (so far) on Solaris 10.


On Thu, Jan 28, 2010 at 11:44 AM, demerphq wrote:

> 2010/1/28 Pas Argenio :
> > Thanks John. Currently, I'm building 32-bit for a Sparc platform and
> also
> > with man CPAN modules such as DBI and DBD.
> >
> > I just rebuilt everything carefully, but still coring often before then
> > displaying a page:
> > Apache 2.2.8 Perl 5.10.0 mod_perl 2.0.3
> >
> > I also turned on perl internal malloc
> >
> > Should I try 64-bit versions next, or fall back to perl 5.8.9?
>
> I recommend you try 5.8.9
>
> Yves
>
>
>
>
>
> --
> perl -Mre=debug -e "/just|another|perl|hacker/"
>

--000e0cd70d0eeb7b4d047e403891
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

I resolved it, finally.=A0 Stupidly, we had created a symbolic link:
=A0=
       libmysqlclient.so.15 -> libmysqlclient.so.16.0.0
<=
br>(I guess our source build of mysql-5.1.32 only provides libmysqlclient.s=
o.16.0.0 but DBD-mysql-4.008 needs libmysqlclient.so.15)

The fix was to get the real libmysqlclient.so.15 library from the mysql.5.0=
..67 Sun package.=A0 We should have known software is rarely FORWARD compati=
ble!
Now to fix it the right way, maybe the latest DBD-mysql uses the .1=
6 library.


Anyway perl 5.10.0 does work (so far) on Solaris 10.


ass=3D"gmail_quote">On Thu, Jan 28, 2010 at 11:44 AM, demerphq "ltr"><><=
/span> wrote:

204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">2010/1/28 Pas Arg=
enio <=
>:


> Thanks John.=A0 Currently, I'm building 32-bit f=
or a Sparc platform and also

> with man CPAN modules such as DBI and DBD.

>

> I just rebuilt everything carefully, but still coring often before the=
n

> displaying a page:

>    Apache 2.2.8   Perl 5.10.0   mod_perl 2.0.3

>

> I also turned on perl internal malloc

>

> Should I try 64-bit versions next, or fall back to perl 5.8.9?



I recommend you try 5.8.9



Yves











--

perl -Mre=3Ddebug -e "/just|another|perl|hacker/"




--000e0cd70d0eeb7b4d047e403891--

Re: mod_perl-2.0.4 with Apache 2.2.9 and perl 5.10.0 intermittent

am 28.01.2010 23:22:37 von Mark Copper

On Thu, Jan 28, 2010 at 04:30:04PM -0500, Pas Argenio wrote:
> I resolved it, finally. Stupidly, we had created a symbolic link:
> libmysqlclient.so.15 -> libmysqlclient.so.16.0.0
> (I guess our source build of mysql-5.1.32 only provides
> libmysqlclient.so.16.0.0 but DBD-mysql-4.008 needs
> libmysqlclient.so.15)
> The fix was to get the real libmysqlclient.so.15 library from the
> mysql.5.0.67 Sun package. We should have known software is rarely
> FORWARD compatible!
> Now to fix it the right way, maybe the latest DBD-mysql uses the .16
> library.
> Anyway perl 5.10.0 does work (so far) on Solaris 10.
>
> On Thu, Jan 28, 2010 at 11:44 AM, demerphq <[1]demerphq@gmail.com>
> wrote:
>
> 2010/1/28 Pas Argenio <[2]pas.argenio@gmail.com>:
>
> > Thanks John. Currently, I'm building 32-bit for a Sparc platform and
> also
> > with man CPAN modules such as DBI and DBD.
> >
> > I just rebuilt everything carefully, but still coring often before
> then
> > displaying a page:
> > Apache 2.2.8 Perl 5.10.0 mod_perl 2.0.3
> >
> > I also turned on perl internal malloc
> >
> > Should I try 64-bit versions next, or fall back to perl 5.8.9?
>
> I recommend you try 5.8.9
> Yves
> --
> perl -Mre=debug -e "/just|another|perl|hacker/"
>
> References
>
> 1. mailto:demerphq@gmail.com
> 2. mailto:pas.argenio@gmail.com
Looks like Bug #37027 for DBD-mysql: mysql_auto_reconnect

I struggled with this one a while back myself.

Mark