Resolved wordpress: Internal error: Decryption failed - Urgent

[finbarr69]

We suffered a severe server crash and I have had to restore parts of /etc /var and /usr from the old hard disk after getting the server re-imaged. For some reason, plesk backups were missing. :-(

Almost back up and running, but when I go into Plesk and under Server Management, I click on Wordpress, I get this error:

Internal error: Decryption failed: error:00000001:lib(0):func(0):reason(1)
Message Decryption failed: error:00000001:lib(0):func(0):reason(1)
File aps_php.php
Line 7173
Type Exception

I have already followed this article, but it has not fixed my problem.

There are two wordpress installations.

In Mysql:
select name,value,registry_object_id from apsc.aps_registry_object_setting where name='admin_password';

gives me :

+----------------+----------------------------------------------------------------+--------------------+
| name | value | registry_object_id |
+----------------+----------------------------------------------------------------+--------------------+
| admin_password | $AES-128-**** | 6 |
| admin_password | $AES-128-**** | 12 |
+----------------+----------------------------------------------------------------+--------------------+

*I have starred out the real passwords*

I assume the passwords I need to update in there were the same as the result from cat /etc/psa/.psa.shadow ? Anyway, I updated them to that and it made no difference.

When I try to visit one of the Wordpress sites I get a 503 Service Temporarily unavailable.

In the error_log. I see:

[proxy:error] [pid 5846:tid 140441563375360] (2)No such file or directory: AH02454: FCGI: attempt to connect to Unix domain socket /var/www/vhosts/system/mydomain.com/php-fpm.sock (*) failed
[proxy_fcgi:error] [pid 5846:tid 140441563375360] AH01079: failed to make connection to backend: httpd-UDS

(I have obfuscated mydomain.com).

I didn't think we used a proxy. Maybe this is part of the problem.

Please, what can I do to get Wordpress up and running again? I've been working all through the night to try to get this resolved. Many thanks!

Brian

 

[finbarr69]

Well! It turned out to be two things. Because I had brought over /etc/passwd but not /etc/shadow ... the passwords for the web users were not on the system. Going into "Web Hosting Access" and re-entering the password for each domain has fixed the websites (no longer a 503 error, but now a normal site!).

Then to fix the Server Management -> Wordpress link (decryption failed message), the page I linked previously did help but it isn't very clear and should really be updated to clarify which password to enter, and it is missing password() to encrypt the password.

I managed to work out that the password set in Web Hosting Access is the one to use in the query like this:

update apsc.aps_registry_object_setting set value=PASSWORD('my_password') where registry_object_id = 6 and name='admin_password';
update apsc.aps_registry_object_setting set value=PASSWORD('my_other_password') where registry_object_id = 12 and name='admin_password';

I found out which one was 6 and which one was 12 by looking at the whole apsc.aps_registry_object_setting table :

select * from apsc.aps_registry_object_setting;

Where you can see the admin_email and title which give you clues about which website is which.

Your registry_object_id values will not be 6 and 12 most likely, but using the full table, you can match things up.

Hope that helps others. now I can go and sleep!