Removing users causes permanent error 503 ;(

[Spacedust]

That's horrible. Even removing /script/restart-web before removing customers does not help.

I'm unable to restart php-fpm because it still listening on some other socket. Killing php-fpm does not help as well.

[Spacedust]

Damn - it's Apache bug !!!

http://www.apache.org/dist/httpd/CHANGES_2.4

Changes with Apache 2.4.13

  *) mod_proxy: Don't put the worker in error state for 500 or 503 errors
     returned by the backend unless failonstatus is configured to.  PR 56925.
     [Yann Ylavic]

[MRatWork]

In design 'restart-web' try to reload but 'restart-web -y' try to 'real' restart. If you want restart-web (without '-y' also execute 'real' restart) for apache, copy /usr/local/lxlabs/kloxo/etc/process/apache to custom.apache and then modified it.

[MRatWork]

Damn - it's Apache bug !!!

http://www.apache.org/dist/httpd/CHANGES_2.4

Changes with Apache 2.4.13

  *) mod_proxy: Don't put the worker in error state for 500 or 503 errors
     returned by the backend unless failonstatus is configured to.  PR 56925.
     [Yann Ylavic]

Just waiting IUS update their httpd24u. I have a plan to use httpd24-httpd (like CentOS 7 do) instead httpd24u.

[Spacedust]

In design 'restart-web' try to reload but 'restart-web -y' try to 'real' restart. If you want restart-web (without '-y' also execute 'real' restart) for apache, copy /usr/local/lxlabs/kloxo/etc/process/apache to custom.apache and then modified it.

I just want reload (graceful) so customers won't see anything, but it crashes because some configs for domains are missing (not yet removed). /script/restart-web should be executed when whole removing processes is ended not after each domain ;(

cPanel has it done someway nobody is seeing anything.

[Spacedust]

Please recompile httpd 2.4.16 in your test repo. We need it.

[MRatWork]

IUS already have 2.4.16 in developing process. I think will be released immediately.

[Spacedust]

IUS already have 2.4.16 in developing process. I think will be released immediately.

Thanks - our customers are having hard times with these 503 errors and we are unable to help them.

[Spacedust]

I got reply for ius:

"Hey Marek,

Thanks for taking the time to reaching out.  Moments before we got your email, we started looking at this new version.  Hopefully we have updated packages in the testing repos soon.


-Ben and the rest of the IUS covedev team"

[Spacedust]

Another reply:

This build is complete and will move to the testing repos during tonight's automation.

Carl George
IUS CoreDev Team

[Spacedust]

Maybe we can set this one ?

https://bz.apache.org/bugzilla/show_bug.cgi?id=56925

"However, if I switch ProxyErrorOverride to Off, the worker would not be put in error state after a request that generates a 500 error, and other clients can still use the same worker."

[Spacedust]

It's released !!!

https://dl.iuscommunity.org/pub/ius/testing/CentOS/6/x86_64/repoview/

[Spacedust]

Got this error again !

[Thu Jul 16 17:04:11 2015] [auth_digest:notice] [pid 19814] [mod_auth_digest.c(236):] [AH01757: generating secret for digest authentication ...] [mail.domena.pl]
[Thu Jul 16 17:04:12 2015] [mpm_event:notice] [pid 19814] [event.c(2807):] [AH00489: Apache/2.4.16 (Red Hat) OpenSSL/1.0.1e-fips configured -- resuming normal operations] [mail.domena.pl]
[Thu Jul 16 17:04:12 2015] [core:notice] [pid 19814] [log.c(1536):] [AH00094: Command line: '/usr/sbin/httpd'] [mail.domena.pl]
[Thu Jul 16 17:04:12 2015] [proxy:error] [pid 19978] [proxy_util.c(2679):] [AH02454: FCGI: attempt to connect to Unix domain socket /opt/configs/php-fpm/sock/apache.sock (127.0.0.1) failed] [default]
[Thu Jul 16 17:04:12 2015] [proxy:error] [pid 19978] [proxy_util.c(2842):] [AH00959: ap_proxy_connect_backend disabling worker for (127.0.0.1) for 60s] [default]
[Thu Jul 16 17:04:12 2015] [proxy_fcgi:error] [pid 19978] [mod_proxy_fcgi.c(924):] [AH01079: failed to make connection to backend: httpd-UDS] [default]
[Thu Jul 16 17:04:13 2015] [proxy:error] [pid 19978] [proxy_util.c(2119):] [AH00940: FCGI: disabled connection for (127.0.0.1)] [default]
[Thu Jul 16 17:04:18 2015] [proxy:error] [pid 19979] [proxy_util.c(2679):] [AH02454: FCGI: attempt to connect to Unix domain socket /opt/configs/php-fpm/sock/apache.sock (127.0.0.1) failed] [default]
[Thu Jul 16 17:04:18 2015] [proxy:error] [pid 19979] [proxy_util.c(2842):] [AH00959: ap_proxy_connect_backend disabling worker for (127.0.0.1) for 60s] [default]
[Thu Jul 16 17:04:18 2015] [proxy_fcgi:error] [pid 19979] [mod_proxy_fcgi.c(924):] [AH01079: failed to make connection to backend: httpd-UDS] [default]

[Spacedust]

It was after updating Ioncube and not restarting the server so maybe it's normal.

[MRatWork]

I am back to use httpd 2.2 for server of this forum because less issues and possible to use fcgid (together suphp) for secondary php.