You are viewing a plain text version of this content. The canonical link for it is here.
Posted to bugs@httpd.apache.org by bu...@apache.org on 2013/12/20 13:35:57 UTC
[Bug 55914] New: child exit segmentation fault after apache graceful
https://issues.apache.org/bugzilla/show_bug.cgi?id=55914
Bug ID: 55914
Summary: child exit segmentation fault after apache graceful
Product: Apache httpd-2
Version: 2.4.7
Hardware: Sun
OS: Solaris
Status: NEW
Severity: regression
Priority: P2
Component: All
Assignee: bugs@httpd.apache.org
Reporter: chriswaddi007@gmail.com
Created attachment 31138
--> https://issues.apache.org/bugzilla/attachment.cgi?id=31138&action=edit
result of pstack on core file generated
hello,
i'm working on Solaris 10 and i've upgraded my apache from 2.2.23 to 2.4.7.
Since then, i have children crashing after a graceful restart. It does not
appear after each graceful restart but sometimes. When it occurs, each time i
have a request in my access log, a child is crashing. Doing a normal restart is
never causing this problem.
Here is a extract of the error log (with LogLevel warn):
[Thu Dec 19 15:51:34.905734 2013] [mpm_worker:notice] [pid 24913:tid 1]
AH00297: SIGUSR1 received. Doing graceful restart
[Thu Dec 19 15:51:36.050587 2013] [mpm_worker:notice] [pid 24913:tid 1]
AH00292: Apache/2.4.7 (Unix) mod_auth_kerb/5.4 OpenSSL/0.9.8k configured --
resuming normal operations
[Thu Dec 19 15:51:36.050644 2013] [core:notice] [pid 24913:tid 1] AH00094:
Command line: '/local/products/sslproxy/bin/httpd'
[Thu Dec 19 15:51:36.050729 2013] [mpm_worker:warn] [pid 24913:tid 1] AH00291:
long lost child came home! (pid 24915)
[Thu Dec 19 15:52:29.076927 2013] [core:notice] [pid 24913:tid 1] AH00052:
child pid 3773 exit signal Segmentation fault (11)
[Thu Dec 19 15:53:29.093903 2013] [core:notice] [pid 24913:tid 1] AH00052:
child pid 3771 exit signal Segmentation fault (11)
[Thu Dec 19 15:54:30.111803 2013] [core:notice] [pid 24913:tid 1] AH00052:
child pid 5049 exit signal Segmentation fault (11)
[Thu Dec 19 15:55:30.129161 2013] [core:notice] [pid 24913:tid 1] AH00052:
child pid 5844 exit signal Segmentation fault (11)
i put the LogLevel to trace8 level to see what's happening just before the
crash:
[Thu Dec 19 18:01:07.144950 2013] [core:trace6] [pid 12540:tid 27]
core_filters.c(527): [client 136.10.10.10:38372] core_output_filter: flushing
because of FLUSH bucket
[Thu Dec 19 18:01:07.147988 2013] [core:trace6] [pid 12540:tid 27]
core_filters.c(527): [client 136.10.10.10:38372] core_output_filter: flushing
because of FLUSH bucket
[Thu Dec 19 18:01:07.164183 2013] [proxy:debug] [pid 12540:tid 27]
proxy_util.c(2035): AH00943: http: has released connection for
(eur-lex.europa.eu)
[Thu Dec 19 18:01:07.164434 2013] [proxy_http:trace2] [pid 12540:tid 27]
mod_proxy_http.c(1793): [client 136.10.10.10:38372] end body send
[Thu Dec 19 18:01:07.200666 2013] [core:notice] [pid 24913:tid 1] AH00051:
child pid 12484 exit signal Segmentation fault (11), possible coredump in
/local/products/sslproxy
[Thu Dec 19 18:01:07.200864 2013] [core:trace4] [pid 24913:tid 1]
mpm_common.c(530): mpm child 12484 (gen 3/slot 1) exited
[Thu Dec 19 18:01:08.209248 2013] [core:trace4] [pid 24913:tid 1]
mpm_common.c(530): mpm child 12545 (gen 3/slot 1) started
And i did a pstack on the core file generated. I will put it in attachment.
Despite the traces put in place, i do not know what is causing this problem.
Can
you help me on this topic? I have to find the root cause otherwise i will have
to go back to apache 2.2.
Thanks for your help .
--
You are receiving this mail because:
You are the assignee for the bug.
---------------------------------------------------------------------
To unsubscribe, e-mail: bugs-unsubscribe@httpd.apache.org
For additional commands, e-mail: bugs-help@httpd.apache.org
[Bug 55914] child exit segmentation fault after apache graceful
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55914
bash@intelerad.com changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |bash@intelerad.com
--
You are receiving this mail because:
You are the assignee for the bug.
---------------------------------------------------------------------
To unsubscribe, e-mail: bugs-unsubscribe@httpd.apache.org
For additional commands, e-mail: bugs-help@httpd.apache.org
[Bug 55914] child exit segmentation fault after apache graceful
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55914
--- Comment #4 from bash@intelerad.com ---
We are experiencing the same issue with httpd-2.4.7. A graceful restart results
in the child segfaulting, in this case it happens to be a mod_jk(1.2.37) child
(if there is such a thing).
We have both mod_logio and mod_status loaded.
We were able to work around the problem by setting the ExtendedStatus directive
to 'off'.
--
You are receiving this mail because:
You are the assignee for the bug.
---------------------------------------------------------------------
To unsubscribe, e-mail: bugs-unsubscribe@httpd.apache.org
For additional commands, e-mail: bugs-help@httpd.apache.org
[Bug 55914] child exit segmentation fault after apache graceful
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55914
Christophe WEIDEMANN <ch...@gmail.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |chriswaddi007@gmail.com
--- Comment #2 from Christophe WEIDEMANN <ch...@gmail.com> ---
What i can tell you is that it appears "randomly" only after a graceful
restart. I did not change any LoadModule items before the restart and i'm not
using mod_logio. The server was not stressed when it happened Concerning the
modules , only standard modules are used except mod_auth_kerb. But i was
already using it with apache 2.2 release without problem.
Details of Loaded Modules:
core_module (static)
so_module (static)
http_module (static)
access_compat_module (shared)
actions_module (shared)
alias_module (shared)
authn_core_module (shared)
auth_basic_module (shared)
auth_kerb_module (shared)
authn_file_module (shared)
authnz_ldap_module (shared)
authz_core_module (shared)
authz_groupfile_module (shared)
authz_host_module (shared)
authz_owner_module (shared)
authz_user_module (shared)
autoindex_module (shared)
socache_shmcb_module (shared)
slotmem_shm_module (shared)
cgid_module (shared)
dir_module (shared)
env_module (shared)
expires_module (shared)
headers_module (shared)
imagemap_module (shared)
include_module (shared)
info_module (shared)
lbmethod_byrequests_module (shared)
ldap_module (shared)
log_config_module (shared)
logio_module (shared)
mime_magic_module (shared)
mpm_worker_module (shared)
mime_module (shared)
negotiation_module (shared)
proxy_module (shared)
proxy_balancer_module (shared)
proxy_connect_module (shared)
proxy_http_module (shared)
rewrite_module (shared)
setenvif_module (shared)
speling_module (shared)
ssl_module (shared)
status_module (shared)
unique_id_module (shared)
unixd_module (shared)
I compiled with the following line:
./configure --prefix=/local/products/revproxy --with-ldap \
--with-ssl=/usr/local/ssl \
--enable-mods-shared=\
"all ssl ldap cache proxy authn_alias file_cache authnz_ldap charset_lite \
dav_lock cache_disk imagemap" \
--with-included-apr \
--enable-nonportable-atomics=yes \
--enable-mpms-shared=all
--
You are receiving this mail because:
You are the assignee for the bug.
---------------------------------------------------------------------
To unsubscribe, e-mail: bugs-unsubscribe@httpd.apache.org
For additional commands, e-mail: bugs-help@httpd.apache.org
[Bug 55914] child exit segmentation fault after apache graceful
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55914
DigitalDude <mi...@gmail.com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |NEEDINFO
--- Comment #5 from DigitalDude <mi...@gmail.com> ---
There is no debug output and no coredumps.
I sometimes get "failed to open stream: HTTP request failed!"
and segmentation fault.
version: httpd-2.4.7
Disabling mod_logio and mod_status with ExtendedStatus=Off does not help in my
case.
--
You are receiving this mail because:
You are the assignee for the bug.
---------------------------------------------------------------------
To unsubscribe, e-mail: bugs-unsubscribe@httpd.apache.org
For additional commands, e-mail: bugs-help@httpd.apache.org
[Bug 55914] child exit segmentation fault after apache graceful
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55914
--- Comment #3 from Rainer Jung <ra...@kippdata.de> ---
(In reply to Christophe WEIDEMANN from comment #2)
> What i can tell you is that it appears "randomly" only after a graceful
> restart. I did not change any LoadModule items before the restart and i'm
> not using mod_logio.
But your below list of loaded modules shows mod_lgio an mod_status, so you
should try whether explicitely setting
ExtendedStatus Off
helps (or disable loading mod_logio and check whether it actually vanished from
the list of loaded modules).
> The server was not stressed when it happened Concerning
> the modules , only standard modules are used except mod_auth_kerb. But i was
> already using it with apache 2.2 release without problem.
Apart from the mod_logio problem I think it should not be the mod_ssl problem I
cited, because you pstack output isn't a good fit for that problem. You might
want to check a few crashes though, whether the pstack results are consistent
(so that it is clear we are looking at the right debug info).
In addition to disable mod_logio or ExtendedStatus Off, suggest you try the
tweo patches I mentioned:
http://svn.apache.org/viewvc?view=revision&revision=1534321
http://svn.apache.org/viewvc?view=revision&revision=1550061
That bug is not about restarts but about a pool lifecycle problem so I can't
rule out that it might also manifest during a restart. Both patches are for
mod_proxy_http and are very small, so easy to apply.
Regards,
Rainer
--
You are receiving this mail because:
You are the assignee for the bug.
---------------------------------------------------------------------
To unsubscribe, e-mail: bugs-unsubscribe@httpd.apache.org
For additional commands, e-mail: bugs-help@httpd.apache.org
[Bug 55914] child exit segmentation fault after apache graceful
Posted by bu...@apache.org.
https://issues.apache.org/bugzilla/show_bug.cgi?id=55914
--- Comment #1 from Rainer Jung <ra...@kippdata.de> ---
Could it be this problem:
https://issues.apache.org/bugzilla/show_bug.cgi?id=50335
In that case the following two patches should help:
http://svn.apache.org/viewvc?view=revision&revision=1534321
http://svn.apache.org/viewvc?view=revision&revision=1550061
Did you change any LoadModule items in configuration before doing the graceful
restart?
Another possible crash I'm aware of is when using both mod_logio and
mod_status. In this case explicitely setting "ExtendedStatus Off" should fix
it.
Are there any non-standard modules active?
Not very likely but also possible:
https://issues.apache.org/bugzilla/show_bug.cgi?id=54357
Regards,
Rainer
--
You are receiving this mail because:
You are the assignee for the bug.
---------------------------------------------------------------------
To unsubscribe, e-mail: bugs-unsubscribe@httpd.apache.org
For additional commands, e-mail: bugs-help@httpd.apache.org