Mailing-List: contact users-help@httpd.apache.org; run by ezmlm
Precedence: bulk
Reply-To: users@httpd.apache.org
Received-SPF: pass (nike.apache.org: domain of ledoc@netcourrier.com
 designates 213.182.54.9 as permitted sender)
From: ledoc@netcourrier.com
To: ph1@openstrike.co.uk,
	users@httpd.apache.org
Date: Fri, 22 Nov 2013 10:05:55 +0100 (CET)
MIME-Version: 1.0
Message-ID: <ea-mime-528f1e73-3a52-6c4075b2@w-4.netcourrier.com>
In-Reply-To: <ea-mime-528e2c66-789e-2398455e@w-3.netcourrier.com>
Content-Type: multipart/alternative;
	boundary="----=_NextPart_001_528f1e73_3a52_4ddf4054"
Subject: Re: [users@httpd] Strange apache behaviour when lauching an
	 externalbinary called by a perl script

------=_NextPart_001_528f1e73_3a52_4ddf4054
Content-Type: text/plain;
	charset="ISO-8859-1"
Content-Transfer-Encoding: quoted-printable

This issue is now resolved: my fortran binary handles very large arrays and=
 complex functions requiring an unlimited stack size.=20

Despite that the stack size limit was defined on a system-wide basis (in /=
etc/security/limits.conf), for any reason it appears that the "apache/perl/=
fortran binary" ensemble was not aware of that (causing my binary to crash =
each time it was called). At the contrary, when I manually restarted apache=
 at the shell prompt, the stacksize limit was correctly passed (.bashrc wit=
h 'ulimit -S -s unlimited').

As a workaround, I used BSD::Resource module (http://search.cpan.org/~jhi/=
BSD-Resource-1.2907/Resource.pm) to define stacksize directly in my perl sc=
ript by using e.g. setrlimit(RLIMIT_STACK, $softlimit, $hardlimit); Thus, t=
hese new stack size limits are now directly passed from my perl script to m=
y binary.=20

Thanks for your time and providing me with insights to fix this issue! Che=
ers.=C2=A0=20


---- Message d'origine ----
De : ledoc@netcourrier.com
=C0 : "Pete Houston" <ph1@openstrike.co.uk>;
=C2=A0=C2=A0=C2=A0users@httpd.apache.org
Objet : Re: Re: [users@httpd] Strange apache behaviour when lauching an ex=
ternalbinary called by a perl script
Date : 21/11/2013 16:53:10 CET

Nice suggestion. But unfortunately both environments are the same (and my =
binary is compiled with static library option).

---- Message d'origine ----
De : "Pete Houston" <ph1@openstrike.co.uk>
=C0 : users@httpd.apache.org
Objet : Re: [users@httpd] Strange apache behaviour when lauching an extern=
albinary called by a perl script
Date : 21/11/2013 14:40:33 CET

Possibly a difference in environment would cause this (perhaps
LD_LIBRARY_PATH or similar?).

Compare the environment in your failing at job with the environment in
your successful shell command and determine if there are any
differences.

HTH,

Pete
--=20
Openstrike - improving business through open source
http://www.openstrike.co.uk/ or call 01722 770036 / 07092 020107

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)

iEYEARECAAYFAlKODVEACgkQdzfnYmsKt524FACgncbrG/UwTOKTXPMp5Zs26FoA
HI4AoKIpwfFsqkJHQSuOdolX0lXnCHWh
=3D6+wF
-----END PGP SIGNATURE-----


------=_NextPart_001_528f1e73_3a52_4ddf4054
Content-Type: text/html;
	charset="ISO-8859-1"
Content-Transfer-Encoding: quoted-printable

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transational//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; charset=3DISO-8859-1=
">
</HEAD>
<BODY>This issue is now resolved: my fortran binary handles very large arr=
ays and complex functions requiring an unlimited stack size. <br><br>Despit=
e that the stack size limit was defined on a system-wide basis (in /etc/sec=
urity/limits.conf), for any reason it appears that the "apache/perl/fortran=
 binary" ensemble was not aware of that (causing my binary to crash each ti=
me it was called). At the contrary, when I manually restarted apache at the=
 shell prompt, the stacksize limit was correctly passed (.bashrc with 'ulim=
it -S -s unlimited').<br><br>As a workaround, I used BSD::Resource module (=
http://search.cpan.org/~jhi/BSD-Resource-1.2907/Resource.pm) to define stac=
ksize directly in my perl script by using e.g. setrlimit(RLIMIT_STACK, $sof=
tlimit, $hardlimit); Thus, these new stack size limits are now directly pas=
sed from my perl script to my binary. <br><br>Thanks for your time and prov=
iding me with insights to fix this issue! Cheers.&nbsp; <br><br><br>---- Me=
ssage d'origine ----<br>
De : ledoc@netcourrier.com<br>
=C0 : "Pete Houston" &lt;ph1@openstrike.co.uk&gt;;<br>
&nbsp;&nbsp;&nbsp;users@httpd.apache.org<br>
Objet : Re: Re: [users@httpd] Strange apache behaviour when lauching an ex=
ternalbinary called by a perl script<br>
Date : 21/11/2013 16:53:10 CET<br>
<br>
Nice suggestion. But unfortunately both environments are the same (and my =
binary is compiled with static library option).<br><br>---- Message d'origi=
ne ----<br>
De : "Pete Houston" &lt;ph1@openstrike.co.uk&gt;<br>
=C0 : users@httpd.apache.org<br>
Objet : Re: [users@httpd] Strange apache behaviour when lauching an extern=
albinary called by a perl script<br>
Date : 21/11/2013 14:40:33 CET<br>
<br>
Possibly a difference in environment would cause this (perhaps<br>
LD_LIBRARY_PATH or similar?).<br>
<br>
Compare the environment in your failing at job with the environment in<br>
your successful shell command and determine if there are any<br>
differences.<br>
<br>
HTH,<br>
<br>
Pete<br>
-- <br>
Openstrike - improving business through open source<br>
http://www.openstrike.co.uk/ or call 01722 770036 / 07092 020107<br>
<br>
-----BEGIN PGP SIGNATURE-----<br>
Version: GnuPG v1.4.11 (GNU/Linux)<br>
<br>
iEYEARECAAYFAlKODVEACgkQdzfnYmsKt524FACgncbrG/UwTOKTXPMp5Zs26FoA<br>
HI4AoKIpwfFsqkJHQSuOdolX0lXnCHWh<br>
=3D6+wF<br>
-----END PGP SIGNATURE-----<br>
<br></BODY></HTML>

------=_NextPart_001_528f1e73_3a52_4ddf4054--