Stefan Fritsch wrote:
On Monday 15 June 2009, Jeroen Hooyberghs wrote:
At each logrotation, the apache proces seg faults and therefor dies
The last message in the /var/log/apache/error.log.1:
[Sun Jun 14 07:41:23 2009] [notice] Graceful restart requested,
doing restart
The first and only message in the error.log:
[Sun Jun 14 07:41:23 2009] [notice] seg fault or similar nasty
error detected in the parent proces
Please try to get a backtrace, as described in
/usr/share/doc/apache2.2-common/README.backtrace
#0 0xb5c9bc80 in ?? ()
No symbol table info available.
#1 0xb76ced8e in CRYPTO_thread_id ()
from /usr/lib/i686/cmov/libcrypto.so.0.9.8
No symbol table info available.
#2 0xb774078c in ERR_get_state () from
/usr/lib/i686/cmov/libcrypto.so.0.9.8
No symbol table info available.
#3 0xb77416b2 in ERR_put_error () from
/usr/lib/i686/cmov/libcrypto.so.0.9.8
No symbol table info available.
#4 0xb772a5d8 in ENGINE_add () from /usr/lib/i686/cmov/libcrypto.so.0.9.8
No symbol table info available.
#5 0xb772ff5a in ENGINE_load_padlock ()
from /usr/lib/i686/cmov/libcrypto.so.0.9.8
No symbol table info available.
#6 0xb772c4c7 in ENGINE_load_builtin_engines ()
from /usr/lib/i686/cmov/libcrypto.so.0.9.8
No symbol table info available.
#7 0xb6d9ab77 in ?? () from /usr/lib/libcurl.so.4
No symbol table info available.
#8 0xb6d9ab69 in ?? () from /usr/lib/libcurl.so.4
No symbol table info available.
#9 0xb6dbe9a0 in ?? () from /usr/lib/libcurl.so.4
No symbol table info available.
#10 0xbf866bc8 in ?? ()
No symbol table info available.
#11 0xb6daef0c in ?? () from /usr/lib/libcurl.so.4
No symbol table info available.
#12 0xb6daeeee in ?? () from /usr/lib/libcurl.so.4
No symbol table info available.
#13 0xb6dbe9a0 in ?? () from /usr/lib/libcurl.so.4
No symbol table info available.
#14 0xbf866bd8 in ?? ()
No symbol table info available.
#15 0xb6da42ad in curl_global_init () from /usr/lib/libcurl.so.4
No symbol table info available.
Backtrace stopped: previous frame inner to this frame (corrupt stack?)
This problem is occuring at each log rotation, but only notice the
problem on 1 debian lenny server I maintain, the other don't seem
to have the issue.
Do you have different modules or different php extensions enabled on
the two servers?
No, I checked on 1 other server that is also serving the same services,
and therefor has exacly the same modules and php extensions enabled.
In any case, as a work-around, you can change /etc/logrotate.d/apache2
to use "restart" instead of "reload".
I changed it for now to do so.
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
