It looks so far like it’s just PHPMyAdmin that broke after the auto-update. The databases still work fine, and can be downloaded from the HestiaCP dashboard.
If PHPMyAdmin alias is blank, then I get a 404 error on PHPMyAdmin, and if I set an alias, then it is a 500 error instead.
I upgraded the server and restarted the server, but no difference. I confirmed that HostName is pointing to the hostname I’m using for PHPMyAdmin.
# phpMyAdmin default Apache configuration
# Alias / /usr/share/phpmyadmin
<Directory /usr/share/phpmyadmin>
Options FollowSymLinks
DirectoryIndex index.php
<IfModule mpm_event_module>
# Use www.conf instead
<FilesMatch \.php$>
SetHandler "proxy:unix:/run/php/www.sock|fcgi://localhost"
</FilesMatch>
</IfModule>
</Directory>
# Disallow web access to directories that don't need it
<Directory /usr/share/phpmyadmin/libraries>
Order Deny,Allow
Deny from All
</Directory>
<Directory /usr/share/phpmyadmin/setup>
Order Deny,Allow
Deny from All
</Directory>
<Directory /usr/share/phpmyadmin/sql>
Order Deny,Allow
Deny from All
</Directory>
<Directory /usr/share/phpmyadmin/vendor>
Order Deny,Allow
Deny from All
</Directory>
<Directory /usr/share/phpmyadmin/templates>
Order Deny,Allow
Deny from All
</Directory>
<Directory /usr/share/phpmyadmin/locale>
Order Deny,Allow
Deny from All
</Directory>
Is it related to how the HestiaCP System PHP version is set to 8.2 but there now appears to be 8.3 and 8.4 on the list? Maybe PHPMyAdmin is trying to use a newer version?
I checked the Apache error log and the warnings on there look the same as the ones that were happening before the update broke PHPMyAdmin
I tried to reinstall PHPMyAdmin by running the /usr/local/hestia/install/upgrade/manual/migrate_phpymyadmin.sh command and it di the following:
[ * ] Backing up old configuration files…
[ * ] Marking phpmyadmin as held in apt…
phpmyadmin set on hold.
[ * ] Removing old folders…
[ * ] Creating new folders…
rm: cannot remove ‘/etc/apache2/conf.d/phpmyadmin.conf’: No such file or directory
[ * ] Installing phpMyAdmin version v5.2.2…
Error: ERROR: Restart of apache2 failed.
[ * ] Creating localhost configuration…
[ * ] Dropping database (could throw an error if successful)…
Then I tried to restart apache2 from the HestiaCP dashboard but it failed to restart. I restarted the server and apache2 is still down and won’t start.
I’m assuming that maybe the HestiaCP auto-update broke PHPMyAdmin in a similar fashion, but now I’m in a bigger predicament since Apache is broken. Is there a reset Hestia command of sorts before I go further down attempting heavier-handed solutions?
It seems that apache2 and nginx are fighting over port 80. When I stop one I can start the other, but when I try to start the second one, it complains the port address is already in use.
I thought that nginx as a proxy server means that apache should be using a different port. Is there not a way to have hestia reconfigure apache properly? I tried doing hestia --reinstall and running the migrate_apache.sh script. Neither script ran into an issue, but neither script resolved the apache port problem (assuming that is the central problem now).
In this process I reinstalled apache2 manually and after restoring the hestia apache conf files, I got the error AH00534: apache2: Configuration error: No MPM loaded., so I did a2enmod mpm_prefork
But now it believes that there’s a syntax error on the apache2.conf line
Order allow,deny
This makes me believe that HestiaCP is using a different version? The reinstalled apache2 is version 2.4.62. Is that not compatible with HestiaCP’s configurations?
Okay, I was able to get apache2 and nginx to play nicely, and I was able to run migrate_phpymyadmin.sh without error, but PHPMyAdmin still shows 404 not found. I restarted nginx and apache and the server and still it still shows not found.