To be fair, the last time I used the mod_php DSO was in 2007 or so. I would
highly recommend to migrate to php-fpm pools instead, and use fcgi to have
httpd communicate with php-fpm.
You'll even gain the ability to do proper privilege separation that way,
and httpd won't have to wait for php
As per this comment [1] I can get httpd to start with systemctl if the
timeout is extended enough. So now I just need to figure out why having PHP
SQLSRV enabled causes so much lag. Thoughts on how to identify the issue?
[1]
`sudo apachectl restart` failure with SQLSRV enabled, success with it
disabled. journalctl output below:
Jul 05 10:38:03 someserver sudo[17255]: someuser : TTY=pts/0 ;
PWD=/etc/rc.d/init.d ; USER=root ; COMMAND=/sbin/apachectl restart
Jul 05 10:38:03 someserver polkitd[879]: Registered
Next step would be to see if using `apachectl restart` works instead of
relying on your distro's init scripts / systemd.
On Tue, Jul 3, 2018 at 11:45 AM James Montalvo
wrote:
> FYI, I posted this question on Server Fault:
> https://serverfault.com/questions/919326/apache-wont-start-with-systemd
FYI, I posted this question on Server Fault:
https://serverfault.com/questions/919326/apache-wont-start-with-systemd
I included in that question the stack traces mentioned in my previous email.
On Mon, Jul 2, 2018 at 2:02 PM James Montalvo
wrote:
> You're right, running `httpd -X` is not dying
You're right, running `httpd -X` is not dying and is properly serving web
pages. It's just dying when running `systemctl (re)start httpd`. It does
not appear to be serving web pages at any point between running `systemctl
start httpd` and when it times out three minutes later.
I ran a stack trace
The last 100 lines of the trace or so should tell you if httpd stopped by
itself because of the mod_php issue. In your log, there is no such
indication.
Try to capture a test case where:
1) That php extension is loaded
2) httpd stops with httpd -X
If httpd keeps running with -X while that
Yes, mod_php. I've run stack traces and didn't see anything stand out. At
the bottom of this email is the last couple hundred lines of the output
from `strace -o /tmp/st -s 5000 httpd -X`, the last couple lines being when
I ctrl-c killed it after several minutes. Do you see anything that I'm
I'm going to venture that you're using the mod_php DSO here. Please clarify
otherwise.
If a mod_php extension prevent httpd from starting normally, I would use
strace to get the exact error, first:
strace -o /tmp/outputfile -s 5000 httpd -X or strace -ff -F -s200 -o
/tmp/strace.out -p PID
It
Hi,
This is my first message to this list. I've reviewed your policies, but my
apologies if I break rules/conventions. Please don't hesitate to correct me.
I've run into an issue where enabling the PHP SQLSRV extension [1] causes
httpd to fail to start. I've documented the issue at length here
10 matches
Mail list logo
