Search results

Please provide additional context. What actions were performed when the error was observed?

Hi, You can install it via PECL. Just go to Tools & Settings > PHP Settings > PHP 8.3 > Manage PECL Packages

What MariaDB version are you using? Older versions are known to allocate RAM and never release it.

Hi, The Nginx Virtual Host looks like this: server { listen x.x.x.x:443 ssl; http2 on; server_name example.net; server_name www.example.net; server_name ipv4.example.net; ssl_certificate /usr/local/psa/var/certificates/***...

You need to set the php settings in that location. Once that location is matched, the other location that handles PHP requests is no longer matched thus any PHP file in /secure will be treated like a plain txt file, exposing the php source rather the generated output. You could enable Proxy Mode...

It sounds like all the database service connection slots get used up. By default, you should have 200 connections and 150 connections per user, so that no single user can use up all the connection slots. Can you run the following when it happens? plesk db "show full processlist"

Plesk repair mail also restarted Dovecot throughout the repair process. That is why the service recovered after running it. A simple systemctl start dovecot would have had the same result. I have noo idea whythe Dovecott service was not enabled and since when.

systemctl enable dovecot

Don't think so. The stored databases don't influence the number of connections. Your service might be stopped, or several users use up all the connection slots.

Your Postfix service might not be set to start on boot. Please share the output of systemctl status postfix If you don't have "service; enabled" run systemctl enable postfix

That repository no longer exists. The upgrade fails because your package manager has a broken repository configured. If you disable the broken repository, the Plesk upgrade will work. Plesk still supports CentOS 7. I don't see any issues with using MariaDB 10.11. MariaDB 11.4 is supported since...

Try plesk bin domain -l. Every subscription can have 1 or more websites.

Disable the repository. Yum doesn't complete any actions because one of the repositories (the MariaDB one) no longer exists. The repositories are in /etc/yum.repos.d/. Find the file that contains the MariaDB URL and change enabled from 1 to 0.

I've just upgraded on my AlmaLinux 8 server and it worked. ===> Plesk database scheme upgrade has been started. Applying migrations from: /usr/local/psa/bootstrapper/pp18.0.69-bootstrapper/migrations/ Migration 18.0.69/2025-04-04-16-09-23_add_id_to_sessions.php applied ===> Plesk database scheme...

Unfortunately, it's not possible to decrypt the Plesk PHP files. Can you select data from the sessions table?

Try to run a Plesk repair: plesk repair db -v Looks like you might have a broken table in the psa database.

Check the headers of emails that are delivered straight to the Trash. It can be a mail filter on the server, check Roundcube settings, or an email client filter. In the headers, you will find the queueid of the email and you can see where it was delivereted by the mail service by searching the...

That looks like a malicious injection. Check if you have a running process spawned by your website system user. First, identify the system user: Next, via SSH, check if there are any running processes: # ps -u example.com_y52t56kk2ld -f UID PID PPID C STIME TTY TIME CMD...

I had to debug such a case recently. The log doesn't show any additional details on why the email client downloaded so much data. You can debug this issue further by enabling mail_debug in Dovecot. Check the value: doveconf | grep mail_debug Edit /etc/dovecot/dovecot.conf removing the # from...

Those requests are normal. It's not a bad bot. What is the IP address on those requests? You can firewall the Plesk port or install the MFA extension for additional protection.