Topic: Apache dies within 1 to 2 minutes of starting. cgroup empty.

[Nigel-Aves]

Wise ones,

Apache version 2.4.6
MySQL version 5.5.52
ISC DHCPd version 4.2.5
Postfix version 2.10.1
PHP 5

I have run into an issue where Apache dies after about 2 mins of starting (even from bootup). This happens in systemctl (and command using Virtualmin and Webmin). As far as I know I have not touched anything to do with Apache,or it's config files. If I do a test of Apache configuration it all runs through with no issues reported.

However, if I run from the command line >> httpd -DFOREGROUND -e debug , Apache runs with no issues at all. I don't have the knowledge to know what the difference is between running httpd -DFOREGROUND and systemctl start httpd.service

I have spend hours and hours on this now trying to find on-line help. To no avail.

Changes I made 2 days ago. Converted Fail2ban to use ipsets and integrated with Shorewall. Followed an on-line tutorial.

Here's what happens, I've removed all the system prompts.


systemctl start httpd
Job for httpd.service failed because a fatal signal was delivered to the control process. See "systemctl status httpd.service" and "journalctl -xe" for details.

First thing to check

systemctl status httpd.service
● httpd.service - The Apache HTTP Server
Loaded: loaded (/usr/lib/systemd/system/httpd.service; enabled; vendor preset: disabled)
Active: failed (Result: signal) since Mon 2017-02-20 16:40:43 MST; 5min ago
Docs: man:httpd(8 )
man:apachectl(8 )
Process: 18576 ExecStart=/usr/sbin/httpd $OPTIONS -DFOREGROUND (code=killed, signal=KILL)
Main PID: 18576 (code=killed, signal=KILL)
CGroup: /system.slice/httpd.service

suexec[18597]: uid: (501/gallery) gid: (501/gallery) cmd: php5.fcgi << Note. Sometimes you only get one line. Once I had none.
suexec[18632]: uid: (506/serendipity) gid: (505/serendipity) cmd: php5.fcgi

Note my server is running 5 web sites. 3 use Piwigo (photo gallery php driven) and two use Joomla. The two mentioned above use Piwigo. Webs site running all OK for 2 years on Centos7.


2nd thing to check. (I've made bold what I believe to be a real clue).

journalctl -xe
---- cut a lot of lines all the same as ones below ------
interface=org.freedesktop.DBus.Properties member=PropertiesChanged
Got message type=signal sender=:1.1 destination=n/a object=/org/freedesktop/systemd1/unit/httpd_2eservice interface=org.freedesktop.DBus.Properties member=PropertiesChanged
Got message type=signal sender=:1.1 destination=n/a object=/org/freedesktop/systemd1/unit/httpd_2eservice interface=org.freedesktop.DBus.Properties member=PropertiesChanged
systemd[1]: Received SIGCHLD from PID 18586 (httpd).
systemd[1]: Child 18586 (httpd) died (code=killed, status=9/KILL)
systemd[1]: Child 18586 belongs to httpd.service
systemd[1]: httpd.service: cgroup is empty
systemd[1]: Sent message type=signal sender=n/a destination=n/a object=/org/freedesktop/systemd1/unit/httpd_2eservice interface=org.freedesktop.DBus.Properties member=PropertiesChanged cookie=47
systemd[1]: Sent message type=signal sender=n/a destination=n/a object=/org/freedesktop/systemd1/unit/httpd_2eservice interface=org.freedesktop.DBus.Properties member=PropertiesChanged cookie=47
systemd-logind[745]: Got message type=signal sender=:1.1 destination=n/a object=/org/freedesktop/systemd1/unit/httpd_2eservice interface=org.freedesktop.DBus.Properties member=PropertiesChanged
systemd-logind[745]: Got message type=signal sender=:1.1 destination=n/a object=/org/freedesktop/systemd1/unit/httpd_2eservice interface=org.freedesktop.DBus.Properties member=PropertiesChanged
systemd[1]: Got cgroup empty notification for: /system.slice/httpd.service
systemd[1]: httpd.service: cgroup is empty
systemd[1]: Sent message type=signal sender=n/a destination=n/a object=/org/freedesktop/systemd1/agent interface=org.freedesktop.systemd1.Agent member=Released cookie=4714 reply_cookie=0 error=n
systemd[1]: Received SIGCHLD from PID 18771 (less).
systemd[1]: Child 18771 (less) died (code=exited, status=2/INVALIDARGUMENT)

Can anyone help?

Many Thanks, Nigel Aves.

[James Blond]

Anything in the apache error log?

[Nigel-Aves]

James,

As it turned out this was operator error!

I had accidently deleted systemd.conf, and this is probably the one thing that is not picked up when running an Apache configuration test! (or mentioned / hint) in any log file.

So it all turned out to be self-inflicted!

Nigel.