I am posting my issue here because my ticket from August 28 has so far remained unanswered and I need an urgent response please (see ticket no. CS16598473).
Here is the problem:
During the night of August 26 to 27, my PLESK server went down. At this point, the exact cause has not yet been identified, but it seems that an attempted update was performed without completing successfully.
I can no longer access the machine at all, whether via SSH or through OVH's SSH (via remote KVM); I cannot execute any restoration command.
When connecting with "Remote KVM", the program window opens, loads for a few seconds and then shows the following error message: "Connection failed!".
When opening with "Serial over LAN (SoL)" the interface displays "[SOL Session operational. Use ~? for help]" and then freezes (unable to type anything).
When connecting to "{my_plesk_server_address}/modules/ssh-terminal/repair/", it loads with the message "Connecting...", then a few seconds later the following error appears:
"Connection Failed
Connection to the plesk-ssh-terminal service failed. Try reconnecting or ask your service provider to make sure the plesk-ssh-terminal service is running."
The fact that remote KVM and the SoL connection fail is a sign that the server is not fully booting up. It’s not a configuration issue; the system is locked or corrupted, and the remote access methods cannot establish a session because the operating system is not responding.
Since the ticket is not progressing, the quickest way is to call OVHcloud telephone support for your region.
It looks like a hardware problem: network, motherboard, power,…
As indicated, test the rescue mode and tell support that the rescue mode also doesn't work. And normally, it will be fixed by replacing the faulty hardware.
Thank you for your help, I will try to launch rescue mode.
I will just have to plan a service interruption because a production website is still running on the PLESK server.
Yes, do it yourself, because you master the shutdown.
Since 2003 I've had OVH servers, because they are very reliable and great at hardware maintenance. What lies on the server afterwards isn’t their problem.
So in such a case, first verify and understand as much as possible and quickly whether the issue is hardware or system, to open the right ticket. A well‑documented hardware ticket triggers a fast intervention and solution, while a too vague ticket can take much longer.
In rescue mode you have access to tools to test memory, disks, which is already part of the investigation; the log files of your server and then you understand everything.
