httpd not restarting from time to time

Lem0nHead · Jan 13, 2006

hello

I'd like to know if it already happened to someone else, and how it was fixed (if was)...
from time to time (once a week or once 2 weeks) my httpd fails on Linux RedHat EL and doesn't restart
when I try to restart it using
# service httpd start
or
# service httpd restart
it just says it was started, but isn't
then I have to do a
# service httpd stop
# killall httpd
# service httpd start

then it starts
probably some childs died, or the main process... don't know... but it's needed to make it restart, otherwise the :80 port will be taken and it won't get to bind on it

my first idea is to check logs, right?
good, but I don't know if it's cause or consequence, but right after that the /etc/httpd/logs/ are (almost) empty and has creation time =~ the time of the error!

my guess:
- /usr/local/bandmin/bandmin (any chances?)
- log rotate (maybe apache gets "confused" and die if a log rotate happens?) - I don't think it's probably... but I wouldn't see why the log file would be empty

any suggestions for finding the cause or fixing it?
or maybe I'll try to "fix" adding a check on (re)start script, making it try to "killall httpd" before starting, but I don't think it would be really a good idea...

thanks

aby · Jan 13, 2006

please do check

/usr/local/apache/logs/error_log for getting more information. Also make sure that apache is enabled in the chkservd
so thatit will automatically restarts if found failed.

Lem0nHead · Jan 14, 2006

I have apache enabled on chkservd, but it doesn't fix it
(my guess: it tries to use the "restart" command, but it doesn't killall before)

here's what I found on /usr/local/apache/logs/error_log

[Sat Jan 14 01:54:58 2006] [notice] Apache/1.3.34 (Unix) mod_auth_passthrough/1.
8 mod_log_bytes/1.2 mod_bwlimited/1.4 PHP/4.4.1 FrontPage/5.0.2.2635 mod_ssl/2.8
.25 OpenSSL/0.9.7a configured -- resuming normal operations
[Sat Jan 14 01:54:58 2006] [notice] suEXEC mechanism enabled (wrapper: /usr/loca
l/apache/bin/suexec)
[Sat Jan 14 01:54:58 2006] [notice] Accept mutex: sysvsem (Default: sysvsem)
[Sat Jan 14 01:55:09 2006] [warn] child process 9549 did not exit, sending anoth
er SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 62 did not exit, sending another
SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 30585 did not exit, sending anot
her SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 28317 did not exit, sending anot
her SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 22804 did not exit, sending anot
her SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 9894 did not exit, sending anoth
er SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 29836 did not exit, sending anot
her SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 15965 did not exit, sending anot
her SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 32616 did not exit, sending anot
her SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 9529 did not exit, sending anoth
er SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 15515 did not exit, sending anot
her SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 21347 did not exit, sending anot
her SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 19683 did not exit, sending anot
her SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 26461 did not exit, sending anot
her SIGHUP
[Sat Jan 14 01:55:09 2006] [warn] child process 10611 did not exit, sending anot
her SIGHUP
[Sat Jan 14 01:55:11 2006] [warn] child process 9549 still did not exit, sending
a SIGTERM
[Sat Jan 14 01:55:15 2006] [error] child process 9549 still did not exit, sendin
g a SIGKILL
[Sat Jan 14 01:55:15 2006] [notice] SIGHUP received. Attempting to restart
[Sat Jan 14 01:55:15 2006] [warn] NameVirtualHost x.x.x.a:80 has no Virtua
lHosts
[Sat Jan 14 01:55:15 2006] [warn] NameVirtualHost x.x.x.b:80 has no Virtua
lHosts
[Sat Jan 14 01:55:15 2006] [warn] NameVirtualHost y.y.y.a:80 has no Virtual
Hosts
[Sat Jan 14 01:55:15 2006] [warn] NameVirtualHost y.y.y.b:80 has no Virtual
Hosts
[Sat Jan 14 01:55:15 2006] [notice] Apache/1.3.34 (Unix) mod_auth_passthrough/1.
8 mod_log_bytes/1.2 mod_bwlimited/1.4 PHP/4.4.1 FrontPage/5.0.2.2635 mod_ssl/2.8
.25 OpenSSL/0.9.7a configured -- resuming normal operations
[Sat Jan 14 01:55:15 2006] [notice] suEXEC mechanism enabled (wrapper: /usr/loca
l/apache/bin/suexec)
[Sat Jan 14 01:55:15 2006] [notice] Accept mutex: sysvsem (Default: sysvsem)
Click to expand...

so looks like it really have problems with childs not exiting...

chirpy · Jan 14, 2006

Indeed. Restart has always been a dubious option with apache because if you have busy child processes it can take some time for apache to exist. It's much kinder, anyway, to restart apache with:

httpd graceful

Unless you have a particular need to kill off httpd entirely, and if you do, then use:

httpd stop

at least twice to ensure that's it's dead, then start it up again with:

httpd startssl

shacker23 · Jun 18, 2007

For the past few months, Apache on my server will not start with "httpd start", not even upon boot. The only way to start Apache is "httpd graceful." "httpd start" does not produce any immediately obvious errors - it acts like it's started correctly, but no httpd processes are launched. This means httpd does not start automatically when the server is rebooted - I always have to start it manually with "graceful."

If I tail the apache error log when running "httpd start," even if apache is not running and there are no httpd threads, I see this:
Code:
[crit] (98)Address already in use: make_sock: could not bind to port 80
But using graceful from a dead stop, I don't get this error and everything starts up normally.

Very weird. Any ideas what could be causing this or how to fix? Thanks.

Forums

The Community Forums

Interact with an entire community of cPanel & WHM users!

httpd not restarting from time to time

Lem0nHead Well-Known Member

aby Well-Known Member

Lem0nHead Well-Known Member

chirpy Well-Known Member Verifed Vendor

shacker23 Well-Known Member

Apache/HTTPD Down

HTTPD shows down in WebHost Manager

httpd service failed to start

httpd.pid not readable after start

The service “httpd” appears to be down.

Share This Page

Useful Searches

The Community Forums

Interact with an entire community of cPanel & WHM users!

httpd not restarting from time to time

Lem0nHead Well-Known Member

aby Well-Known Member

Lem0nHead Well-Known Member

chirpy Well-Known Member Verifed Vendor

shacker23 Well-Known Member

Apache/HTTPD Down

HTTPD shows down in WebHost Manager

httpd service failed to start

httpd.pid not readable after start

The service “httpd” appears to be down.

Share This Page