Starting Apache as service quits with SIGWINCH

Question

When running service apache start I see in the log file this entry:

[mpm_event:notice] [pid 1906:tid XXX] AH00489: Apache/2.4.23 (Unix)
                         OpenSSL/1.0.2g PHP/7.0.9 configured -- resuming normal operations
[core:notice] [pid 1906:tid XXX] AH00094: Command line: '/usr/sbin/httpd'
[mpm_event:notice] [pid 1906:tid XXX] AH00492: caught SIGWINCH, shutting down gracefully

and the service apache does not start? I can find in internet what SIGWINCH means [Window size change] but it does not really help me in this case.

The file /etc/systemd/system/apache.service:

[Unit]
Description=The Apache Webserver [FaF Compiled]
After=network.target nss-lookup.target time-sync.target
[email protected] plymouth-quit.service xdm.service

[Service]
Type=notify
PrivateTmp=true
ExecStart=/usr/sbin/httpd -k start
ExecReload=/usr/sbin/httpd -k graceful
ExecStop=/usr/sbin/httpd -k graceful-stop
KillMode=mixed

[Install]
WantedBy=multi-user.target
Alias=httpd.service apache.service

Running on SLES 12 SP1. I have compiled Apache on my own. Starting it with apachectl -k start works perfectly and I can access and run the PHP code.

My question: What am I doing wrong that the apache service does not start. I have enabled it with systemctl enable apache. This was the output:

ln -s '/usr/lib/systemd/system/apache2.service' '/etc/systemd/system/httpd.service'
ln -s '/usr/lib/systemd/system/apache2.service' '/etc/systemd/system/apache.service'
ln -s '/usr/lib/systemd/system/apache2.service' '/etc/systemd/system/multi-user.target.wants/apache2.service'

How were `SIGWINCH` made fatal? The default action is to discard the signal. — thrig, Aug 08 '16 at 17:01

score 3 · Answer 1 · answered Feb 03 '18 at 12:27

systemd is instructing Apache to stop with graceful-stop, which generates a SIGWINCH signal, and hence the log caught SIGWINCH, shutting down gracefully. (the SIGWINCH signal is (ab)used by Apache)

IMO, there is a possibility of another error, including configuration errors, making it stop, and you are looking at related symptoms which are normal behaviour.

I would recommend perusing your Apache error logs, usually by default in SLES under the directory /var/log/apache2.

score 1 · Answer 2 · edited Jun 11 '20 at 14:16

1

I found a similar problem reported on RedHat. Closed with a status "NOTABUG". Eventually recognised as wrong config.

img #1

img #2

img #3

UPDATE

I found one more similar problem reported on serverfault. And one more in the answer to it, in the image below.

img 1/1

edited Jun 11 '20 at 14:16

Community

1

answered Aug 08 '16 at 20:05

`-DFOREGROUND` does not help in my case. Apache does start but it does not return back to `service` and therefore `service` complains with `apache2.service start operation timed out. Terminating.` – Peter VARGA Aug 08 '16 at 20:18
Did you read the thread in the link? – Aug 08 '16 at 20:19
1

I gave up. I understand the situation but I don't have a clue how to fix it. I found out when I remove `ExecReload=` from `/usr/lib/systemd/system/apache2.service` then `SIGWINCH` is not sent but then Apache does not start - no error message in log, nothing. It just immediately quits when it has been started. I implemented now the hack that I run `apachectl -k start` as a cron job when the system starts. I can restart Apache with `apachectl -k restart`. I will try it later; may be I get some new ideas. – Peter VARGA Aug 08 '16 at 22:16
I found one more, different. See update. – Aug 08 '16 at 22:37

score 1 · Answer 3 · edited Mar 05 '19 at 14:54

1

From a comment on the Red Hat bug report linked from Tomasz's answer:

It looks like you need to run 00-systemd.conf if you want systemd to recognize that httpd has started successfully.

I ran this and it worked:

ln -sr /etc/httpd/conf.modules.d/00-systemd.conf /etc/httpd/conf.d/00-systemd.conf
systemctl restart httpd

I'm on Amazon Linux 2 (will probably also work on RHEL/Centos 7).

edited Mar 05 '19 at 14:54

Anthony Geoghegan

12,605
7
59
62

answered Nov 14 '18 at 20:03

luds

111
1

Thanks for giving credit, but, … where did you get this? – G-Man Says 'Reinstate Monica' Nov 14 '18 at 20:43
@G-Man I've edited this answer to correctly attribute that quote. – Anthony Geoghegan Mar 05 '19 at 14:55

score 0 · Answer 4 · answered May 08 '20 at 00:48

0

This problem is from adding too many domains at once.
You are supposed to add one domain then restart httpd. I made this mistake too.

answered May 08 '20 at 00:48

powerhousemethod

435
9
25

Could you please elaborate a little on _why_ this fixes the problem? – Martin Konrad May 08 '20 at 03:25
I hate apache so much if this is the issue. I just have 7 vHosts. – Robert J Apr 29 '22 at 04:39

Rob Myrick · Answer 5 · 2023-01-01T18:52:41.310

0

I had this issue happening in AWS Cloud9 Console. I thought that SIGWINCH by itself was the problem, but later found a default setting in AWS Cloud 9 that terminates the instance after 30 minutes. This triggers SIGWINCH, which then tells Apache to shut down.

I had no idea about this, and this took me 2 days to figure out. I hope it saves someone time.

edited Jan 01 '23 at 18:52

answered Jan 01 '23 at 06:42

Rob Myrick

101
2

Your answer could be improved with additional supporting information. Please [edit] to add further details, such as citations or documentation, so that others can confirm that your answer is correct. You can find more information on how to write good answers [in the help center](/help/how-to-answer). – Community Jan 02 '23 at 21:16

Starting Apache as service quits with SIGWINCH

5 Answers5

UPDATE