Hello I'm running CentOS 5.3 with httpd-2.2.3-22.el5.centos.x86_64 and php-5.1.6-23.2.el5_3.x86_64. When the logrotate scripts run and send the HUP signal to httpd, the httpd process quits instead of reloading. The only thing I can find in the logs is this: [Sun Apr 19 04:02:04 2009] [notice] seg fault or similar nasty error detected in the parent process There wasn't any segfault reported in dmesg and I couldn't find anything else in the other logs (messages, etc...) It's reproducable every time, when I send the HUP signal to the master httpd process the same thing happens. My configuration is nothing special, just the stock one, with a few unnecesary modules commented out and one virtual host. Thanks, Glenn
>----- Oorspronkelijk bericht ----- >Van: RedShift [mailto:redshift at pandora.be]>Verzonden: zondag , april 19, 2009 09:42 AM>Aan: centos at centos.org>Onderwerp: [CentOS] httpd crashes after signal HUP> >Hello > > >I'm running CentOS 5.3 with httpd-2.2.3-22.el5.centos.x86_64 and php-5.1.6-23.2.el5_3.x86_64. When the logrotate scripts run and send the HUP signal to httpd, the httpd process quits instead of reloading. The only thing I can find in the logs is this: > >[Sun Apr 19 04:02:04 2009] [notice] seg fault or similar nasty error detected in the parent process > >There wasn't any segfault reported in dmesg and I couldn't find anything else in the other logs (messages, etc...) > >It's reproducable every time, when I send the HUP signal to the master httpd process the same thing happens. > >My configuration is nothing special, just the stock one, with a few unnecesary modules commented out and one virtual host. >A follow-up, I cannot seem to reproduce this on my laptop (using i386 instead of x86_64) with the same configuration. Glenn
Have you tried this: http://www.electrictoolbox.com/seg-fault-or-similar-nasty-error/ Per --- Original message follows --- SUBJECT:?Re: [CentOS] httpd crashes after signal HUP FROM: ?"RedShift" TO:?"CentOS mailing list" DATE:?19-04-2009 9:49>----- Oorspronkelijk bericht ----- >Van: RedShift [mailto:redshift at pandora.be]>Verzonden: zondag , april 19, 2009 09:42 AM>Aan: centos at centos.org>Onderwerp: [CentOS] httpd crashes after signal HUP>>Hello > > >I'm running CentOS 5.3 with httpd-2.2.3-22.el5.centos.x86_64 andphp-5.1.6-23.2.el5_3.x86_64. When the logrotate scripts run and send the HUP signal to httpd, the httpd process quits instead of reloading. The only thing I can find in the logs is this:> >[Sun Apr 19 04:02:04 2009] [notice] seg fault or similar nasty errordetected in the parent process> >There wasn't any segfault reported in dmesg and I couldn't findanything else in the other logs (messages, etc...)> >It's reproducable every time, when I send the HUP signal to themaster httpd process the same thing happens.> >My configuration is nothing special, just the stock one, with a fewunnecesary modules commented out and one virtual host.>A follow-up, I cannot seem to reproduce this on my laptop (using i386 instead of x86_64) with the same configuration. Glenn _______________________________________________ CentOS mailing list CentOS at centos.org http://lists.centos.org/mailman/listinfo/centos -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.centos.org/pipermail/centos/attachments/20090419/0150ee75/attachment-0002.html>
Per Qvindesland wrote:> Have you tried this: > http://www.electrictoolbox.com/seg-fault-or-similar-nasty-error/ > > PerYes, I have tried it, but that's a workaround, not a real solution. Glenn> > --- Original message follows --- > > *Subject: *Re: [CentOS] httpd crashes after signal HUP > *From: *"RedShift" <redshift at pandora.be> > *To: *"CentOS mailing list" <centos at centos.org> > *Date: *19-04-2009 9:49 > > > >----- Oorspronkelijk bericht ----- > >Van > : RedShift [mailto:redshift at pandora.be] > >Verzonden > : zondag > , april > 19, 2009 09:42 AM > >Aan > : centos at centos.org > >Onderwerp > : [CentOS] httpd crashes after signal HUP > > > >Hello > > > > > >I'm running CentOS 5.3 with httpd-2.2.3-22.el5.centos.x86_64 and > php-5.1.6-23.2.el5_3.x86_64. When the logrotate scripts run and send > the HUP signal to httpd, the httpd process quits instead of > reloading. The only thing I can find in the logs is this: > > > >[Sun Apr 19 04:02:04 2009] [notice] seg fault or similar nasty > error detected in the parent process > > > >There wasn't any segfault reported in dmesg and I couldn't find > anything else in the other logs (messages, etc...) > > > >It's reproducable every time, when I send the HUP signal to the > master httpd process the same thing happens. > > > >My configuration is nothing special, just the stock one, with a > few unnecesary modules commented out and one virtual host. > > > > A follow-up, I cannot seem to reproduce this on my laptop (using > i386 instead of x86_64) with the same configuration. > > > Glenn > >
RedShift wrote:> Hello > > > I'm running CentOS 5.3 with httpd-2.2.3-22.el5.centos.x86_64 and php-5.1.6-23.2.el5_3.x86_64. When the logrotate scripts run and send the HUP signal to httpd, the httpd process quits instead of reloading. The only thing I can find in the logs is this: > > [Sun Apr 19 04:02:04 2009] [notice] seg fault or similar nasty error detected in the parent process > > There wasn't any segfault reported in dmesg and I couldn't find anything else in the other logs (messages, etc...) > > It's reproducable every time, when I send the HUP signal to the master httpd process the same thing happens. > > My configuration is nothing special, just the stock one, with a few unnecesary modules commented out and one virtual host. > > > Thanks, > > > Glenn > >I figured out it was coming from a PHP extension called MAPI, that comes shipped with a piece of software I use (Zarafa). This problem didn't occur on my laptop because it didn't have the mapi.so extension loaded. So this is not a CentOS problem. For reference: http://forums.zarafa.com/viewtopic.php?f=9&t=2225 Best regards, Glenn