Description of problem: systemctl restart httpd.service hangs. Abortable via CTRL-C Version-Release number of selected component (if applicable): httpd-2.4.3-12.fc18.i686 How reproducible: Always Steps to Reproduce: 1. On a Fedora 18 x86 computer, yum install httpd-2.4.3-12.fc18.i686 2. systemctl start httpd.service 3. systemctl restart httpd.service Alternative: 3. apachectl graceful Actual results: [philippe@victor ~]$ sudo systemctl restart httpd.service ^C[philippe@victor ~]$ ps -ef | grep httpd apache 6267 1 0 16:16 ? 00:00:00 /usr/sbin/httpd -DFOREGROUND philippe 6856 21252 0 16:29 pts/1 00:00:00 grep --color=auto httpd root 20945 20872 0 08:59 pts/0 00:00:00 sudo tail -f /var/log/httpd/access_log root 20947 20945 0 08:59 pts/0 00:00:00 tail -f /var/log/httpd/access_log [philippe@victor ~]$ sudo systemctl status httpd.service httpd.service - The Apache HTTP Server Loaded: loaded (/usr/lib/systemd/system/httpd.service; enabled) Active: deactivating (final-sigterm) since Sun 2013-02-10 16:28:15 CET; 1min 10s ago Process: 6820 ExecStop=/usr/sbin/httpd $OPTIONS -k graceful-stop (code=exited, status=0/SUCCESS) Main PID: 8547 (code=exited, status=0/SUCCESS) Status: "Total requests: 2555; Current requests/sec: 0; Current traffic: 0 B/sec" CGroup: name=systemd:/system/httpd.service `-6267 /usr/sbin/httpd -DFOREGROUND Feb 09 22:06:19 victor.vouters.dyndns.org httpd[8547]: [Sat Feb 09 22:06:19.0... Feb 09 22:06:19 victor.vouters.dyndns.org httpd[8547]: [Sat Feb 09 22:06:19.0... Feb 09 22:06:24 victor.vouters.dyndns.org systemd[1]: Started The Apache HTTP... Feb 10 16:28:15 victor.vouters.dyndns.org systemd[1]: Stopping The Apache HTT... Feb 10 16:28:16 victor.vouters.dyndns.org httpd[6820]: [Sun Feb 10 16:28:16.9... Feb 10 16:28:16 victor.vouters.dyndns.org httpd[6820]: [Sun Feb 10 16:28:16.9... [philippe@victor ~]$ ps -ef | grep httpd apache 6267 1 0 16:16 ? 00:00:00 /usr/sbin/httpd -DFOREGROUND philippe 6879 21252 0 16:29 pts/1 00:00:00 grep --color=auto httpd root 20945 20872 0 08:59 pts/0 00:00:00 sudo tail -f /var/log/httpd/access_log root 20947 20945 0 08:59 pts/0 00:00:00 tail -f /var/log/httpd/access_log [philippe@victor ~]$ sudo kill -TERM httpd [philippe@victor ~]$ ps -ef | grep httpd philippe 6921 21252 0 16:30 pts/1 00:00:00 grep --color=auto httpd root 20945 20872 0 08:59 pts/0 00:00:00 sudo tail -f /var/log/httpd/access_log root 20947 20945 0 08:59 pts/0 00:00:00 tail -f /var/log/httpd/access_log [philippe@victor ~]$ sudo systemctl start httpd.service [philippe@victor ~]$ ps -ef | grep httpd root 6931 1 8 16:30 ? 00:00:00 /usr/sbin/httpd -DFOREGROUND apache 6934 6931 0 16:30 ? 00:00:00 /usr/sbin/httpd -DFOREGROUND apache 6935 6931 0 16:30 ? 00:00:00 /usr/sbin/httpd -DFOREGROUND apache 6936 6931 0 16:30 ? 00:00:00 /usr/sbin/httpd -DFOREGROUND apache 6937 6931 0 16:30 ? 00:00:00 /usr/sbin/httpd -DFOREGROUND apache 6938 6931 0 16:30 ? 00:00:00 /usr/sbin/httpd -DFOREGROUND philippe 6941 21252 0 16:30 pts/1 00:00:00 grep --color=auto httpd root 20945 20872 0 08:59 pts/0 00:00:00 sudo tail -f /var/log/httpd/access_log root 20947 20945 0 08:59 pts/0 00:00:00 tail -f /var/log/httpd/access_log Notice the CTRL-C on the first command and the apa&he process left existing. Expected results: Clean stop/start of the httpd daemon Additional info:
BTW: /usr/sbin/httpd -DFOREGROUND this is also wrong, as demons do not start in foreground mode by default. foreground blocks the start process.
fix it with a change of inside your unit file https://bugzilla.redhat.com/show_bug.cgi?id=907579 That will fork apache into background, like it always should do :)
"customercare" - you are providing workarounds rather than fixes. Philippe, could you say: a) What modules you have loaded? (PHP etc) b) Provide the output of "tail /var/log/httpd/error_log"
You have to correct /lib/systemd/system/httpd.service which is NOT used with apachectl -k xxxxxx [philippe@victor ~]$ sudo apachectl -k start Passing arguments to httpd using apachectl is no longer supported. You can only start/stop/restart httpd using this script. If you want to pass extra arguments to httpd, edit the /etc/sysconfig/httpd config file. [philippe@victor ~]$ ps -ef | grep httpd root 2468 1 4 00:24 ? 00:00:00 /usr/sbin/httpd -k start apache 2471 2468 0 00:24 ? 00:00:00 /usr/sbin/httpd -k start apache 2472 2468 0 00:24 ? 00:00:00 /usr/sbin/httpd -k start apache 2473 2468 0 00:24 ? 00:00:00 /usr/sbin/httpd -k start apache 2474 2468 0 00:24 ? 00:00:00 /usr/sbin/httpd -k start apache 2475 2468 0 00:24 ? 00:00:00 /usr/sbin/httpd -k start philippe 2478 29942 0 00:24 pts/1 00:00:00 grep --color=auto httpd root 23238 1706 0 Feb23 pts/0 00:00:00 sudo tail -f /var/log/httpd/access_log root 23240 23238 0 Feb23 pts/0 00:00:00 tail -f /var/log/httpd/access_log [philippe@victor ~]$ sudo apachectl -k stop Passing arguments to httpd using apachectl is no longer supported. You can only start/stop/restart httpd using this script. If you want to pass extra arguments to httpd, edit the /etc/sysconfig/httpd config file. [philippe@victor ~]$ ps -ef | grep http philippe 2620 29942 0 00:28 pts/1 00:00:00 grep --color=auto http root 23238 1706 0 Feb23 pts/0 00:00:00 sudo tail -f /var/log/httpd/access_log root 23240 23238 0 Feb23 pts/0 00:00:00 tail -f /var/log/httpd/access_log philippe 29917 1 0 Feb18 ? 00:00:00 /usr/libexec/gvfsd-http --spawner :1.3 /org/gtk/gvfs/exec_spaw/1
Here is what needs to be coded in /lib/systemd/system/httpd.service: [philippe@victor ~]$ sudo cat /lib/systemd/system/httpd.service [Unit] Description=The Apache HTTP Server After=network.target remote-fs.target nss-lookup.target [Service] Type=forking EnvironmentFile=/etc/sysconfig/httpd ExecStart=/usr/sbin/httpd $OPTIONS -k start #ExecStart=/usr/sbin/httpd $OPTIONS -DFOREGROUND ExecReload=/usr/sbin/httpd $OPTIONS -k graceful ExecStop=/usr/sbin/httpd $OPTIONS -k graceful-stop PrivateTmp=true [Install] WantedBy=multi-user.target [philippe@victor ~]$ With this coding: [philippe@victor ~]$ sudo systemctl --system daemon-reload [philippe@victor ~]$ ps -ef | grep http root 2034 1873 0 10:47 pts/1 00:00:00 sudo tail -f /var/log/httpd/access_log root 2036 2034 0 10:47 pts/1 00:00:00 tail -f /var/log/httpd/access_log philippe 5996 4234 0 12:05 pts/0 00:00:00 grep --color=auto http [philippe@victor ~]$ sudo systemctl start httpd.service [philippe@victor ~]$ ps -ef | grep http root 2034 1873 0 10:47 pts/1 00:00:00 sudo tail -f /var/log/httpd/access_log root 2036 2034 0 10:47 pts/1 00:00:00 tail -f /var/log/httpd/access_log root 6004 1 7 12:05 ? 00:00:00 /usr/sbin/httpd -k start apache 6006 6004 0 12:05 ? 00:00:00 /usr/sbin/httpd -k start apache 6007 6004 0 12:05 ? 00:00:00 /usr/sbin/httpd -k start apache 6008 6004 0 12:05 ? 00:00:00 /usr/sbin/httpd -k start apache 6009 6004 0 12:05 ? 00:00:00 /usr/sbin/httpd -k start apache 6010 6004 0 12:05 ? 00:00:00 /usr/sbin/httpd -k start philippe 6013 4234 0 12:05 pts/0 00:00:00 grep --color=auto http [philippe@victor ~]$ sudo systemctl stop httpd.service [philippe@victor ~]$ ps -ef | grep http root 2034 1873 0 10:47 pts/1 00:00:00 sudo tail -f /var/log/httpd/access_log root 2036 2034 0 10:47 pts/1 00:00:00 tail -f /var/log/httpd/access_log philippe 6185 4234 0 12:09 pts/0 00:00:00 grep --color=auto http [philippe@victor ~]$ sudo systemctl stop httpd.service [philippe@victor ~]$ ps -ef | grep http root 2034 1873 0 10:47 pts/1 00:00:00 sudo tail -f /var/log/httpd/access_log root 2036 2034 0 10:47 pts/1 00:00:00 tail -f /var/log/httpd/access_log philippe 6185 4234 0 12:09 pts/0 00:00:00 grep --color=auto http [philippe@victor ~]$ sudo systemctl restart httpd.service [philippe@victor ~]$ ps -ef | grep http root 2034 1873 0 10:47 pts/1 00:00:00 sudo tail -f /var/log/httpd/access_log root 2036 2034 0 10:47 pts/1 00:00:00 tail -f /var/log/httpd/access_log root 6227 1 4 12:10 ? 00:00:00 /usr/sbin/httpd -k start apache 6230 6227 0 12:10 ? 00:00:00 /usr/sbin/httpd -k start apache 6231 6227 0 12:10 ? 00:00:00 /usr/sbin/httpd -k start apache 6232 6227 0 12:10 ? 00:00:00 /usr/sbin/httpd -k start apache 6233 6227 0 12:10 ? 00:00:00 /usr/sbin/httpd -k start apache 6234 6227 0 12:10 ? 00:00:00 /usr/sbin/httpd -k start philippe 6237 4234 0 12:10 pts/0 00:00:00 grep --color=auto http [philippe@victor ~]$ sudo systemctl status httpd.service httpd.service - The Apache HTTP Server Loaded: loaded (/usr/lib/systemd/system/httpd.service; enabled) Active: active (running) since Sun 2013-02-24 12:10:15 CET; 5min ago Process: 6223 ExecStop=/usr/sbin/httpd $OPTIONS -k graceful-stop (code=exited, status=0/SUCCESS) Process: 6226 ExecStart=/usr/sbin/httpd $OPTIONS -k start (code=exited, status=0/SUCCESS) Main PID: 6227 (httpd) Status: "Total requests: 0; Current requests/sec: 0; Current traffic: 0 B/sec" CGroup: name=systemd:/system/httpd.service |-6227 /usr/sbin/httpd -k start |-6230 /usr/sbin/httpd -k start |-6231 /usr/sbin/httpd -k start |-6232 /usr/sbin/httpd -k start |-6233 /usr/sbin/httpd -k start `-6234 /usr/sbin/httpd -k start Feb 24 12:10:13 victor.vouters.dyndns.org systemd[1]: Starting The Apache HTT... Feb 24 12:10:15 victor.vouters.dyndns.org systemd[1]: Started The Apache HTTP... [philippe@victor ~]$
Dear Customercare, My only way to get $ sudo apachectl {start|stop|restart|graceful| and so on} work like before as well as $ sudo systemctl {start|stop|restart} httpd.service has been to modify /lib/systemd/system/httpd.service this way [philippe@victor C]$ sudo cat /lib/systemd/system/httpd.service [Unit] Description=The Apache HTTP Server After=network.target remote-fs.target nss-lookup.target [Service] Type=forking EnvironmentFile=/etc/sysconfig/httpd ExecStart=/usr/sbin/httpd $OPTIONS -k start #ExecStart=/usr/sbin/httpd $OPTIONS -DFOREGROUND ExecReload=/usr/sbin/httpd $OPTIONS -k graceful ExecStop=/usr/sbin/httpd $OPTIONS -k graceful-stop PrivateTmp=true [Install] WantedBy=multi-user.target [philippe@victor C]$ This is exactly what I document in the corresponding Red Hat's bugzilla report. Grüsse von Frankreich
Dear Red Hat httpd package maintainer, I want and need to inform you that my change to /lib/systemd/system/httpd.service although seemingly working is not perfect. I have not correctly isolated the actual problem. One direction I am inclined to look at is that Apache 2.4 would create over the time additional child processes (at least one in my case) which would be out of the knowledge of systemctl (so out of its control). Stay tuned awaiting for more information. Best regards, Philippe Vouters (Fontainebleau/France)
Dear Red Hat httpd package maintainer, The direction where I was inclined to look at (read previous post) which could explain why sudo systemctl stop httpd.service could hang does not look obvious. I do notice that unlike what I was usually doing my two actions are stop followed by start instead of a single restart. [philippe@victor ~]$ ps -ef | grep "/usr/sbin/httpd -k start" | wc -l 8 [philippe@victor ~]$ sudo systemctl status httpd.service | grep "/usr/sbin/httpd -k start" | wc -l 7 [philippe@victor ~]$ sudo systemctl stop httpd.service [philippe@victor ~]$ sudo systemctl start httpd.service [philippe@victor ~]$ Stay tuned for more input. Best regards, Philippe Vouters (Fontainebleau/France)
Rather looks like there is a race condition within systemctl with it not being reliable enough. So far I have only been able to notice it with the Apache service. [philippe@victor ~]$ sudo vi /etc/httpd/conf.d/tima.conf [philippe@victor ~]$ sudo systemctl restart httpd.service [philippe@victor ~]$ sudo vi /etc/sysconfig/iptables [philippe@victor ~]$ sudo systemctl restart httpd.service (HANGED) ^C [philippe@victor ~]ps -ef | grep "/usr/sbin/httpd -k start" | wc -l 2 [philippe@victor ~]$ ps -ef | grep "/usr/sbin/httpd -k start" apache 26458 1 0 13:19 ? 00:00:00 /usr/sbin/httpd -k start philippe 26643 23574 0 13:23 pts/1 00:00:00 grep --color=auto /usr/sbin/httpd -k start [philippe@victor ~]$ sudo systemctl status httpd.service | grep "/usr/sbin/httpd -k start" 26656 /usr/sbin/httpd -k start 26658 /usr/sbin/httpd -k start 26659 /usr/sbin/httpd -k start 26660 /usr/sbin/httpd -k start 26661 /usr/sbin/httpd -k start 26662 /usr/sbin/httpd -k start [philippe@victor ~]$ sudo systemctl restart iptables.service [philippe@victor ~]$ sudo systemctl status httpd.service httpd.service - The Apache HTTP Server Loaded: loaded (/usr/lib/systemd/system/httpd.service; enabled) Active: active (running) since Fri 2013-04-26 13:23:39 CEST; 8min ago Process: 26583 ExecStop=/usr/sbin/httpd $OPTIONS -k graceful-stop (code=exited, status=0/SUCCESS) Process: 26654 ExecStart=/usr/sbin/httpd $OPTIONS -k start (code=exited, status=0/SUCCESS) Main PID: 26656 (/usr/sbin/httpd) CGroup: name=systemd:/system/httpd.service 26656 /usr/sbin/httpd -k start 26658 /usr/sbin/httpd -k start 26659 /usr/sbin/httpd -k start 26660 /usr/sbin/httpd -k start 26661 /usr/sbin/httpd -k start 26662 /usr/sbin/httpd -k start 26926 /usr/sbin/httpd -k start Apr 26 13:23:38 victor.vouters.dyndns.org systemd[1]: Starting The Apache HTT... Apr 26 13:23:39 victor.vouters.dyndns.org systemd[1]: Started The Apache HTTP... [philippe@victor ~]$ Interestingly and despite the fact that, unlike previously, I did not take any action to manually kill -9 the remaining Apache process, I now read: [philippe@victor ~]$ ps -ef | grep "/usr/sbin/httpd -k start" root 26656 1 0 13:23 ? 00:00:00 /usr/sbin/httpd -k start apache 26658 26656 0 13:23 ? 00:00:00 /usr/sbin/httpd -k start apache 26659 26656 0 13:23 ? 00:00:00 /usr/sbin/httpd -k start apache 26660 26656 0 13:23 ? 00:00:00 /usr/sbin/httpd -k start apache 26661 26656 0 13:23 ? 00:00:00 /usr/sbin/httpd -k start apache 26662 26656 0 13:23 ? 00:00:00 /usr/sbin/httpd -k start apache 26926 26656 0 13:27 ? 00:00:00 /usr/sbin/httpd -k start apache 27481 26656 0 13:39 ? 00:00:00 /usr/sbin/httpd -k start philippe 27569 23574 0 13:41 pts/1 00:00:00 grep --color=auto /usr/sbin/httpd -k start AND THE SURVIVING APACHE PROCESS DISAPPEARED: [philippe@victor ~]$ ps -ef | grep "/usr/sbin/httpd -k start" | grep 26458 [philippe@victor ~]$ So next time I shall NOT CRTL-C '$ sudo systemctl restart httpd.service' if it appears to hang and observe things from another terminal. The actual problem definition may be something totally other. A likelyhood can be Apache too slow to graceful-stop. Stay tuned. Warmest regards, Philippe Vouters (Fontainebleau/France)
Created attachment 740381 [details] System Monitor Screenshot
Oops ! I lost what I entered. The truth I now come up with is that $ sudo systemctl restart httpd.service appears to hang but it does NOT. The software action reveals abnormally terribly slow with no visible reason in CPU/Memory/Swap/Network activity (see System Monitor Screenshot). I document http://vouters.dyndns.org/tima/All-OS-Apache-Invalidating_Apache_server_cache.html on my Web site with the htcacheclean action I took at Apache 2.2 times. I can't view my disk activity (the disk LED does NOT work on my desktop as well as my disk making no noise as well as it being a MAXTOR SATA disk with a memory backup cache proving far much faster than my old ATA Seagate disk). My next step will thus be to redo the htcacheclean action or, better, shall dig over Apache 2.4 Web help to completely invalidate any Apache disk cache. Will it improve the systemctl restart httpd.service response time ? Another possible direction where I can look at is a Apache lock not correctly and fast enough unlocked. Stay tuned. Philippe
Did the following change inside my Web site Apache configuration file: The more so as I read: [philippe@victor ~]$ sudo lsof | grep httpd | more | grep mod_cache lsof: WARNING: can't stat() fuse.gvfsd-fuse file system /run/user/500/gvfs Output information may be incomplete. /usr/sbin 31293 root mem REG 8,4 31312 4559390 /usr/lib/httpd/modules/mod_cache_disk.so /usr/sbin 31293 root mem REG 8,4 64256 4555408 /usr/lib/httpd/modules/mod_cache.so /usr/sbin 31295 apache mem REG 8,4 31312 4559390 /usr/lib/httpd/modules/mod_cache_disk.so /usr/sbin 31295 apache mem REG 8,4 64256 4555408 /usr/lib/httpd/modules/mod_cache.so /usr/sbin 31296 apache mem REG 8,4 31312 4559390 /usr/lib/httpd/modules/mod_cache_disk.so /usr/sbin 31296 apache mem REG 8,4 64256 4555408 /usr/lib/httpd/modules/mod_cache.so /usr/sbin 31297 apache mem REG 8,4 31312 4559390 /usr/lib/httpd/modules/mod_cache_disk.so /usr/sbin 31297 apache mem REG 8,4 64256 4555408 /usr/lib/httpd/modules/mod_cache.so /usr/sbin 31298 apache mem REG 8,4 31312 4559390 /usr/lib/httpd/modules/mod_cache_disk.so /usr/sbin 31298 apache mem REG 8,4 64256 4555408 /usr/lib/httpd/modules/mod_cache.so /usr/sbin 31299 apache mem REG 8,4 31312 4559390 /usr/lib/httpd/modules/mod_cache_disk.so /usr/sbin 31299 apache mem REG 8,4 64256 4555408 /usr/lib/httpd/modules/mod_cache.so /usr/sbin 31459 apache mem REG 8,4 31312 4559390 /usr/lib/httpd/modules/mod_cache_disk.so /usr/sbin 31459 apache mem REG 8,4 64256 4555408 /usr/lib/httpd/modules/mod_cache.so [philippe@victor ~]$ Here my unique Web site Apache configuration change: # # Disable the cache lock # <IfModule mod_cache.c> CacheLock off CacheLockPath /tmp/mod_cache-lock CacheLockMaxAge 5 </IfModule> This is suggested by http://httpd.apache.org/docs/current/mod/mod_cache.html I tried several successive sudo systemctl restart httpd.service and could not incur any abnormal long delay to return to the bash prompt. To be further confirmed over time. Stay tuned. Philippe
Looks there is a bug inside httpd or in libc (the int_free call retrieved twice on the call stack). httpd is linked with libpthread.so.0. A likelyhood is that int_free (libc calls in general ???) is not signal-safe. This is showed by this fact on the gdb backtrace below: "#8 <signal handler called>" and the call trace which preceedes. This reminds me my fix proposal to aterm (https://bugzilla.redhat.com/show_bug.cgi?id=835908). If my fix proposal to aterm is what should be done, the bug is in Apache 2.4 forgetting to protect itself against some signal(s) to be determined. During a hanging systemctl restart and the knowledge of the httpd pid, I have been doing: [philippe@victor ~]$ sudo strace -v -e trace=all -f -o apache.log -p 1048 [philippe@victor ~]$ sudo cat apache.log 1048 futex(0xb73a4420, FUTEX_WAIT_PRIVATE, 2, NULL <unfinished ...> 1048 +++ killed by SIGKILL +++ and $ sudo gdb /usr/sbin/httpd -p 1558 (gdb) where #0 0xb76bf424 in __kernel_vsyscall () #1 0xb72b67d2 in __lll_lock_wait_private () from /lib/libc.so.6 #2 0xb722de93 in _L_lock_4410 () from /lib/libc.so.6 #3 0xb7228d79 in _int_free () from /lib/libc.so.6 #4 0xb739f7ae in apr_allocator_destroy () from /lib/libapr-1.so.0 #5 0xb73a015c in apr_pool_destroy () from /lib/libapr-1.so.0 #6 0xb6ff17cf in clean_child_exit () from /etc/httpd/modules/mod_mpm_prefork.so #7 0xb6ff218b in just_die () from /etc/httpd/modules/mod_mpm_prefork.so #8 <signal handler called> #9 0xb7228b7f in _int_free () from /lib/libc.so.6 #10 0xb68f6456 in zend_hash_destroy () from /etc/httpd/modules/libphp5.so #11 0xb68ddd50 in destroy_zend_class () from /etc/httpd/modules/libphp5.so #12 0xb68f643e in zend_hash_destroy () from /etc/httpd/modules/libphp5.so #13 0xb68e8147 in zend_shutdown () from /etc/httpd/modules/libphp5.so #14 0xb6883262 in php_module_shutdown () from /etc/httpd/modules/libphp5.so #15 0xb6883345 in php_module_shutdown_wrapper () from /etc/httpd/modules/libphp5.so #16 0xb699a18f in php_apache_child_shutdown () from /etc/httpd/modules/libphp5.so #17 0xb739ffbe in apr_pool_destroy () from /lib/libapr-1.so.0 #18 0xb6ff17cf in clean_child_exit () from /etc/httpd/modules/mod_mpm_prefork.so before the process gets killed.
This last crash looks like this fixed Bug 912288 fixed in httpd-2.4.4-4.fc19.
Unless I misunderstood, this fix ought to be in systemctl. "fix service file to not send SIGTERM after ExecStop (#906321, #912288)" This would mean other services might also be impacted. I would need to test this specific fix built for FC18 i686. How can I manage to get this accurate fix WITHOUT modifying ANY other code parts ? This way I shall be able to best confirm this SIGTERM sending by systemctl is the actual root cause of the problems I encountered. Perhaps the other FC18 X86_64 user who reported a heap httpd glibc crash on systemctl stop httpd.service may also confirm. Also I notice that /sbin/apachectl is very bound to /usr/bin/systemctl activations (start|stop|restart|status|graceful|graceful-stop).
Provided systemctl was allowing kill -TERM|HUP|<other> `cat /var/run/pid` there would be less problems. [philippe@victor ~] # these httpd processes have started manually with [philippe@victor ~] # sudo httpd -k start, [philippe@victor ~]$ ps -ef | grep httpd philippe 9362 23363 0 21:07 pts/1 00:00:00 grep --color=auto httpd philippe 10480 1775 0 Apr28 pts/0 00:00:00 tail -f /var/log/httpd/access_log root 12498 1 0 14:02 ? 00:00:00 /sbin/httpd -k start apache 12501 12498 0 14:02 ? 00:00:00 /sbin/httpd -k start apache 12502 12498 0 14:02 ? 00:00:00 /sbin/httpd -k start apache 12503 12498 0 14:02 ? 00:00:00 /sbin/httpd -k start apache 12504 12498 0 14:02 ? 00:00:00 /sbin/httpd -k start apache 12505 12498 0 14:02 ? 00:00:00 /sbin/httpd -k start apache 12707 12498 0 14:07 ? 00:00:00 /sbin/httpd -k start apache 15291 12498 0 15:13 ? 00:00:00 /sbin/httpd -k start apache 15621 12498 0 15:22 ? 00:00:00 /sbin/httpd -k start apache 15623 12498 0 15:22 ? 00:00:00 /sbin/httpd -k start apache 15624 12498 0 15:22 ? 00:00:00 /sbin/httpd -k start [philippe@victor ~]$ sudo kill -TERM 12498 [philippe@victor ~]$ ps -ef | grep httpd philippe 9378 23363 0 21:07 pts/1 00:00:00 grep --color=auto httpd philippe 10480 1775 0 Apr28 pts/0 00:00:00 tail -f /var/log/httpd/access_log [philippe@victor ~]$ sudo systemctl start httpd.service [philippe@victor ~]$ ps -ef | grep httpd root 9403 1 0 21:08 ? 00:00:00 /usr/sbin/httpd -k start apache 9406 9403 0 21:08 ? 00:00:00 /usr/sbin/httpd -k start apache 9407 9403 0 21:08 ? 00:00:00 /usr/sbin/httpd -k start apache 9408 9403 0 21:08 ? 00:00:00 /usr/sbin/httpd -k start apache 9409 9403 0 21:08 ? 00:00:00 /usr/sbin/httpd -k start apache 9410 9403 0 21:08 ? 00:00:00 /usr/sbin/httpd -k start apache 9411 9403 0 21:08 ? 00:00:00 /usr/sbin/httpd -k start philippe 9609 23363 0 21:10 pts/1 00:00:00 grep --color=auto httpd philippe 10480 1775 0 Apr28 pts/0 00:00:00 tail -f /var/log/httpd/access_log [philippe@victor ~]$ sudo kill -TERM 9403 [philippe@victor ~]$ ps -ef | grep httpd philippe 9624 23363 0 21:10 pts/1 00:00:00 grep --color=auto httpd philippe 10480 1775 0 Apr28 pts/0 00:00:00 tail -f /var/log/httpd/access_log [philippe@victor ~]$ sudo systemctl status httpd.service httpd.service - The Apache HTTP Server Loaded: loaded (/usr/lib/systemd/system/httpd.service; enabled) Active: inactive (dead) since Mon 2013-04-29 21:10:44 CEST; 20s ago Process: 9619 ExecStop=/usr/sbin/httpd $OPTIONS -k graceful-stop (code=exited, status=0/SUCCESS) Process: 9401 ExecStart=/usr/sbin/httpd $OPTIONS -k start (code=exited, status=0/SUCCESS) Main PID: 9403 (code=exited, status=0/SUCCESS) CGroup: name=systemd:/system/httpd.service Apr 29 21:08:14 victor.vouters.dyndns.org systemd[1]: Started The Apache HTTP... Apr 29 21:10:44 victor.vouters.dyndns.org httpd[9619]: httpd (no pid file) no... [philippe@victor ~]$ sudo systemctl start httpd.service [philippe@victor ~]$ sudo systemctl status httpd.service httpd.service - The Apache HTTP Server Loaded: loaded (/usr/lib/systemd/system/httpd.service; enabled) Active: active (running) since Mon 2013-04-29 21:11:50 CEST; 26s ago Process: 9619 ExecStop=/usr/sbin/httpd $OPTIONS -k graceful-stop (code=exited, status=0/SUCCESS) Process: 9669 ExecStart=/usr/sbin/httpd $OPTIONS -k start (code=exited, status=0/SUCCESS) Main PID: 9670 (/usr/sbin/httpd) CGroup: name=systemd:/system/httpd.service 9670 /usr/sbin/httpd -k start 9673 /usr/sbin/httpd -k start 9674 /usr/sbin/httpd -k start 9675 /usr/sbin/httpd -k start 9676 /usr/sbin/httpd -k start 9677 /usr/sbin/httpd -k start Apr 29 21:11:49 victor.vouters.dyndns.org systemd[1]: Starting The Apache HTT... Apr 29 21:11:50 victor.vouters.dyndns.org systemd[1]: Started The Apache HTTP...
Well, the fix was confirmed in Bug 912288 and packages are already in F18 repository. The symptoms and also the backtrace are the same as in mentioned bug, so I'm closing this as duplicate. If you find out your problem persists, feel free to reopen. *** This bug has been marked as a duplicate of bug 912288 ***