Plesk 12.5 RTM is out

[cjhmdm]

After upgrading to 12.5 (wish we'd known it was in RTM status and not stable... the installer should provide a warning for this), and now none of our websites work (they're all 404)

Initially, I ran into this error:

AH00526: Syntax error on line 1 of /etc/apache2/conf-enabled/php_cgi_wrapper.conf:
Invalid command 'scriptAlias', perhaps misspelled or defined by a module not included in the server configuration

Warning : failed to perform deferred action : httpd_reconfigure
 Trying to restart service apache2... Restarting web server: apache2 failed!
There are processes named 'apache2' running which do not match your pid file which are left untouched in the name of safety, Please review the situation by hand. ... (warning).

In which I had to manually enable the alias module

Then, apache couldn't bind to port 80, so I had to manually disable the nginx reverse proxy in order to get apache to start/restart.

From there, all of our sites are returning 404 errors, even after having successfully regenerated our apache configs.
Did a little searching and ran the following as well (both successfully) to no avail:
/usr/local/psa/bootstrapper/pp12.5.30-bootstrapper/bootstrapper.sh repair
/usr/local/psa/admin/bin/httpdmng --reconfigure-all

No errors for either, but again after restarting apache2, all of my sites return 404.

So at this point, our entire server is essentially down since none of our websites can be accessed. So what do we do now?

 

[IgorG]

So what do we do now?

Try

# plesk repair web

 

[cjhmdm]

Try

# plesk repair web

Tried that already as well, no errors and no dice:

# plesk repair web

Checking Plesk version .............................................. [OK]

Checking for custom configuration templates ......................... [OK]

Checking for the JkWorkersFile directive in the Apache configuration  [OK]

Checking associations between domains and IP addresses .............. [OK]

Checking for corrupted reference between IP collections and IP
addresses ........................................................... [OK]

Checking for links between APS applications and subscriptions ....... [OK]

Checking for the Zend extension declaraion in php.ini ............... [OK]

Check symbolic links for latest virtual host config files ........... [OK]

Checking for system users home directories consistency .............. [OK]

Checking for records with empty name field in the Configurations table[OK]

Checking for nginx ULIMIT value ..................................... [OK]

Checking for extra configurations in database not owned by any object [OK]

Repairing web server configuration

  Reinstall SSL certificates and set the default SSL certificate for all IP addresses? [Y/n]
    Reinstalling SSL certificates ................................... [OK]
    Applying the default SSL certificate to all IP addresses ........ [OK]

  Repair server-wide configuration parameters for web servers? [Y/n]
    Repairing server-wide configuration parameters for web servers .. [OK]

  Update the file of sharing passwords and permissions of users according to actual information? [Y/n]
    Updating the file of sharing passwords and permissions of users
    according to actual information ................................. [OK]

  Repair web server configuration for all domains? [Y/n]
    Repairing web server configuration for all domains .............. [OK]

Checking the usage of PHP handlers .................................. [OK]

Error messages: 0; Warnings: 0; Errors resolved: 0

Edit: I should also note that disabling/enabling nginx has no bearing on the issue I'm facing.

 

[UFHH01]

There are processes named 'apache2' running which do not match your pid file which are left untouched in the name of safety, Please review the situation by hand. ... (warning).

If you see such errors, please be reminded, that you have to kill all the orphaned apache processes, before you can restart apache again with the new configuration.

ps aux | grep apache
kill -9 XXXX ( where XXXX is/are the number(s) of the pidfile(s) of your running processes )

restart apache after your kills and post possible errors, please.​

 

[cjhmdm]

If you see such errors, please be reminded, that you have to kill all the orphaned apache processes, before you can restart apache again with the new configuration.

ps aux | grep apache
kill -9 XXXX ( where XXXX is/are the number(s) of the pidfile(s) of your running processes )

restart apache after your kills and post possible errors, please.​

Hello, yes, I'm past this error. The cause of that error was due to nginx already running on port 80.

 

[cjhmdm]

Ok, so I've been able to restore one of the sites (I'm only working on 1 at a time) by changing the following settings:
Enable Nginx
Domain > PHP Settings > change FROM FastCGI application served by Apache TO FPM application served by Nginx (NONE of the other php settings work)
Domain > Apache & nginx settings > tick both "Smart static files processing" and "Serve static files directly by nginx"

So, in a nutshell, I have to run everything through nginx because running anything through Apache causes 404 errors (for both php and static files).

This wouldn't be that big of an issue for me since I use nginx as my webserver on my non plesk machines but we have 100's of domains on this machine and having to go through and change the plesk setting as well as creating nginx rewrites for each one individually isn't practical.

So I'm back to square one really, except that I now know the problem lies with apache.

Anyone have any ideas that I haven't already tried?

Edit: It really just seems like apache is ignoring the conf files because as far as I can tell from years of doing vhosts manually, they are correct. All files for each domain are also where they are supposed to be.

So here is the list of symptoms thus far:
Upgrading from 11.5.x OR 12.0.x to 12.5.x results in the following error:

AH00526: Syntax error on line 1 of /etc/apache2/conf-enabled/php_cgi_wrapper.conf:
Invalid command 'scriptAlias', perhaps misspelled or defined by a module not included in the server configuration

Warning : failed to perform deferred action : httpd_reconfigure
Trying to restart service apache2... Restarting web server: apache2 failed!
There are processes named 'apache2' running which do not match your pid file which are left untouched in the name of safety, Please review the situation by hand. ... (warning).

In which case you have to do 2 things to resolve the issue:
1. edit /etc/apache2/conf-enabled/php_cgi_wrapper.conf and change "scriptAlias" to "ScriptAlias"
2. Enable mod alias in apache settings (or via a2enmod).

After resolving that issue and regenerating conf files, you then have to stop nginx from a terminal (ssh) before you can restart apache. Stopping nginx via plesk control panel does nothing at this point.

After that comes the real problems:

1.Apache serves no files at all. As stated above, it's as if the vhost conf files are being completely ignored, even though they appear to be included correctly.

2. It doesn't matter what wrapper for php is used, the only one that works is php fpm through nginx

3. Static files only work when "Serve static files directly by nginx" is ticked.

4. Doing it this way is both counter-intuitive as well as non-practical as several more steps would need to be taken manually on a per domain basis to get a fully functional website (eg. creating all the necessary nginx rules for say, wordpress).


At this point in time, I am completely stuck, and losing money because I cannot access my websites.

 

[UFHH01]

Hi cjhmdm,

to solve your very specific issue(s), I recommend to open a new thread, with your PHP issues and Plesk version as a title, defining as well your OS and other possible informations ( apache version, nginx version, domain specific php5-fpm configuration files, .... ), which might help to post specific solutions for you.


First, I would check, that your system PHP - settings are correct and working as expected for apache and nginx. it is now advantageous, to use at least PHP version 5.4. Depending to your system, upgrade your PHP version with apt-get or yum. If you need help with that, please define your OS and the current PHP - version.


Please be reminded, that the following structure overrules each previous setting for PHP:

If you know that, then you could change some standard settings over the service plan and synchronize the service plan afterwards, to take effect to the depending domains ( previous domain specific settings will be overwritten ) You have as well the choice to do that over the command line:

List PHP - settings for a service plan:

/usr/local/psa/bin/service_plan --show-php-settings <service-plan-name>​

To add a custom PHP - setting, as for example "open_basedir = {WEBSPACEROOT}{/}{:}{TMP}{/}{:}{/}usr{/}share{/}php{/}{:}{/}usr{/}share{/}php5{/}{:}{/}usr{/}share{/}pear{/}" or/and "error_reporting = E_ALL & ~E_DEPRECATED & ~E_NOTICE & E_USER_DEPRECATED & E_USER_WARNING", please create a file ( YOUR_DEFINITION_FILE_NAME ) with the following content:

open_basedir = {WEBSPACEROOT}{/}{:}{TMP}{/}{:}{/}usr{/}share{/}php{/}{:}{/}usr{/}share{/}php5{/}{:}{/}usr{/}share{/}pear{/}
error_reporting = E_ALL & ~E_DEPRECATED & ~E_NOTICE & E_USER_DEPRECATED & E_USER_WARNING

... and execute the command:

/usr/local/psa/bin/service_plan --update-php-settings <service-plan-name> -settings YOUR_DEFINITION_FILE_NAME​

Update PHP - settings for service plan:

/usr/local/psa/bin/service_plan --update <service-plan-name> -php true -php_handler_id fastcgi
​

Please see all possible CLI commands for the "service_plan" utility at:

Hosting Plans: service_plan Utility ( Online documentation: Command Line Reference, Plesk 12.5 for Linux )​

... or use the command:

/usr/local/psa/bin/service_plan --help​

If you need further help, please consider a separate thread for it.​

After upgrading to 12.5 (wish we'd known it was in RTM status and not stable... the installer should provide a warning for this), and now none of our websites work (they're all 404)

Stabled versions are declared as "( stable )", when you use autoinstaller over the command line.

 

[hardbrasil]

Hello Igor,

is possible to use the old Template on 12.5?
i am not worried about responsive, i didn't like the new visual.

good job!

 

[Necroman]

Hi! I'm running Plesk 12.0.18 #63 on CentOS 6.7 and enabled Early Adopters release tier (Tools & Settings > Update & Upgrade Settings) but when i go to updates the upgrade to Plesk 12.5 doesn't show. Do i have to wait a while or am i missing something?

From what i read, this isn't a stable version yet? I know the meaning of RTM, so this is the version the resellers are currently testing? When can we expect a stable and final version? Thanks

 

[hardbrasil]

Hi! I'm running Plesk 12.0.18 #63 on CentOS 6.7 and enabled Early Adopters release tier (Tools & Settings > Update & Upgrade Settings) but when i go to updates the upgrade to Plesk 12.5 doesn't show. Do i have to wait a while or am i missing something?

From what i read, this isn't a stable version yet? I know the meaning of RTM, so this is the version the resellers are currently testing? When can we expect a stable and final version? Thanks

Hi, for me works just changing to "Early Adopters release".
the options to upgrade will be under "Tools & Settings > Update & Upgrade Settings".


now i am checking another plesk and its not available like you said, maybe them disable for a while.

 

[UFHH01]

Hi Necroman,

When can we expect a stable and final version? Thanks

Please view the answers at: http://download1.parallels.com/Plesk/PP12/parallels-plesk-12.5-for-linux-change-log.html ( Comment section )

"Early Adopters Release" is there to test and to fix potential issues/problems, to make sure that the "Final" ( stable ) release can be used on all supported systems. A possible un-available source of a "Early Adopters Release" should be a temporary issue or a mirror - issue. Please try an hour later and you ( mostly ) will have no issue(s) at all to reach the source as expected.

 

[Josh12]

This bug linked below was reported over a year ago and still hasn't been fixed. Makes making changes to Nginx more difficult then they should be. Can we see it fixed in 12.5 final?
http://kb.odin.com/en/122689

 

[IgorG]

This bug linked below was reported over a year ago and still hasn't been fixed. Makes making changes to Nginx more difficult then they should be. Can we see it fixed in 12.5 final?
http://kb.odin.com/en/122689

If you mean bug PPPM-1914 it is still under developer's investigation and fix planned for next release.

 

[IgorG]

Hello Igor,

is possible to use the old Template on 12.5?
i am not worried about responsive, i didn't like the new visual.

good job!

If you go to Extensions -> Skins and Color Shemes you will see big button - Revert to 12.0 Color Scheme.

 

[gregconway]

Hi,

I have upgraded 3 x CentOS 6 servers to Plesk 12.5.30 and I have the following issue with all three of them since the upgrade.

When sending mail from the command-line I now see this message -

plesk sendmail[6080]: sendmail argv = send-mail -i -- [email protected]
send-mail: warning: /etc/postfix/main.cf, line 713: overriding earlier entry: mailbox_size_limit=0
postdrop: warning: /etc/postfix/main.cf, line 713: overriding earlier entry: mailbox_size_limit=0

The mail does arrive at the destination but never used to see this message before the update.

Thanks,

Greg.

 

[danami]

Some cron bugs I'm seeing now on all our 12.5.30 servers:

/etc/cron.daily/0logwatch:
plesk sendmail[4011]: sendmail argv = sendmail -t

Also:

/etc/cron.hourly/plesk-php-cleanuper:
Warning: readlink failed on /proc/20582/fd/645 : No such file or directory

 

[gregconway]

Just to confirm that I am also seeing this error on CentOS 5 / Plesk 12.5.30, having updated another server!

plesk sendmail[27539]: sendmail argv = send-mail -i [email protected]

(when sending mail from the command line)

Thanks,

Greg.

 

[IgorG]

sendmail argv issue has been discussed already here http://talk.plesk.com/threads/ppp-1...untu-sendmail-notifications-on-screen.334229/

 

[Alban Staehli]

Hi,
I wish to clarify the support aspect of the "Early Adopters" releases.
IF I update to Plesk 12.5, are you still able to support me for any technical issues? I'm speaking about paid support, as for any other stable release.
Thx for your feedback.

 

[IgorG]

Hi,
I wish to clarify the support aspect of the "Early Adopters" releases.
IF I update to Plesk 12.5, are you still able to support me for any technical issues? I'm speaking about paid support, as for any other stable release.
Thx for your feedback.

Yes, be sure that you will be fully supported by Odin Support Team.