At night, the server stops responding, in kvm you can see Unit systemd-journald.service entered failed state, how to find the reason?

Question

linux

Greetings!
At night, after 12 hours, the server stops responding to requests, packets do not reach it by ping, and you can’t connect to the shell either.
There is kvm, connecting through it, you can see the endless output of such lines:

Unit systemd-journald.service entered failed state.
systemd-journald.service has no holdoff time, schelduling restart.
Stopping Journal Service...
Starting Journal Service...
Started Journal Service.
systemd-journald.service watchdog timeout (limit 1min)!

It is treated only by rebooting through the same kvm. Didn't find anything clear on google.
The server is running a medium-load site (LAMP), several docker containers.
What could be the problem? What logs to look at to understand what the trouble is?
UPD:
This output in /var/log/messages is clear during a crash:
/var/log/messages (yadisk)
Syslog output from crash to reboot
/var/log/syslog (yadisk)
ATOP continues to work even during crash

atop output 10 minutes before crash

atop output 1 minute after crash

atop output 10 minutes after crash

Answer 1

For some reason, the magazine got stuck. Do you have disk space?