Xenforo not working for me with php7.0 - what have I missed?

Mouth

Well-known member
Seeking some thoughts/advise on what I've missed when trying to get my XF dev site working with php7.0

php7.0 is working fine. I have files of test.php and test2.php in my XF directory, with
Code:
phpinfo();
and
Code:
var_export($_SERVER)
both working and outputting as expected.

Trying to run XF via / or /admin.php is failing though, with an error of
Code:
2016/03/05 15:25:32 [error] 39996#39996: *43 recv() failed (104: Connection reset by peer) while reading response header from upstream, client: <redacted>, server: mysite.com, request: "GET /admin.php HTTP/2.0", upstream: "fastcgi://unix:/run/php/php7.0-fpm.sock:", host: "mysite.com"

Not sure why php7.0 is slamming the door on XF, whilst allowing test files to run OK?

I also run xenforo_requirements.php in the XF directory and it returns no issues.

Any suggestions on what to test further?
 
Hello,

Slow scripts are killed after a said timeout because it is configured that way.
With other words scripts taking longer than a specified time would get killed and nginx would report a recv or readv error as the connection is closed from the php-fpm engine/process.

Kind regards,
George.
 
Slow scripts are killed after a said timeout because it is configured that way.
With other words scripts taking longer than a specified time would get killed and nginx would report a recv or readv error as the connection is closed from the php-fpm engine/process.
Thanks. My fpm/pool/www.conf is set with ...
Code:
; The timeout for serving a single request after which a PHP backtrace will be
; dumped to the 'slowlog' file. A value of '0s' means 'off'.
; Available units: s(econds)(default), m(inutes), h(ours), or d(ays)
; Default Value: 0
;request_slowlog_timeout = 0
... so I don't think its a slow timeout being triggered. The error is also returned instantly when executing XF (/ or admin.php)
 
When I execute XF (/ or admin.php) I note that pool www exits and restarts ...
Code:
[05-Mar-2016 16:00:58] WARNING: [pool www] child 48706 exited on signal 11 (SIGSEGV) after 1.470827 seconds from start
[05-Mar-2016 16:00:58] NOTICE: [pool www] child 48719 started
... it doesn't do this when running the test scripts.
 
Last edited:
Thanks. My fpm/pool/www.conf is set with ...
Code:
; The timeout for serving a single request after which a PHP backtrace will be
; dumped to the 'slowlog' file. A value of '0s' means 'off'.
; Available units: s(econds)(default), m(inutes), h(ours), or d(ays)
; Default Value: 0
;request_slowlog_timeout = 0
... so I don't think its a slow timeout being triggered. The error is also returned instantly when executing XF (/ or admin.php)

That configures the log. With the ; in front of the line that line is not set so its at the same value being 0. All this means is the log is going to be empty. You can enable the log to find out what scripts are a problem specifically.

The actual PHP timeout is another value altogether.
 
Are you running ANY third party addons?

Also what does view source show on the blank page? (if anything)
 
Last edited:
Are you running ANY third party addons?
Also what does view source show on the blank page? (if anything)
Yes, but I have "$config['enableListeners'] = 0;" within my config.php to remove that as a possible cause.
I get the default nginx error page.
 
Additionally, last night I purged and re-installed php7.0. The only config change I made was to point to the mysql.sock file so that mysql connections work. Still same result - phpinfo(); and
xenforo_requirements.php etc. works but xf does not.
 
What happens if you run the requirements test script that can be downloaded here?

https://xenforo.com/purchase/

That will tell you if xenforo can run with your php/mysql version. It is worth a try.

Editing my original post. Sorry sir, I did not read carefully your previous post where you mentioned that you have tried that script.
 
Resolved :)

For any else that may be searching for a solution to the same/similar issue, I had to add
memcached.serializer = php

to my php7.0 config.

Appears that memcached and igbinary are still unstable.
 
Top Bottom