Finding the causes of a server crash on FreeBSD

Question

System administration

The server worked and worked and suddenly stopped responding (http, ping, ssh, ftp - no results).
We called the hosting company. They said that the server was frozen, now it has been rebooted and now everything is working again.

Everything ended well (they rebooted promptly), but I would like to understand where to look for the reasons for the fall?

In /val/log/messages, the last ones were several unsuccessful authorizations from root via ssh (a standard thing - bots are constantly trying to guess the password to root, which is forbidden) and the following: after this, entries after a reboot in /var/log/dmesg. today is also nothing interesting. Can you please tell me where else can I find something useful? PS Based on the results, you could make a short How To :)

Feb 26 16:09:57 client monit[71718]: monit: Socket 5 close failed -- Connection reset by peer
Feb 26 16:52:11 client kernel: mfi0: 31687 (352053149s/0x0020/info) - Patrol Read complete


Answer 1

last issues left logins?
next time ask for a screen of the dependency - the dependency of the dependency is different.
If the swap is gone, you need to look for the cause of excessive memory consumption. Often this is Apache, whose maxclients are set far beyond the physical capabilities of the server.
I have a frya on a couple of servers for some reason lately it has often spontaneously rebooted ...

Answer 2

What platform?
We have a storm on the irq ride on the Intel platform. They have not been able to diagnose yet because the machine is runtime and somehow works ... In addition, even with a high load on the network card, the network falls off. And the machine is sometimes overloaded from a high load. Everything is fine with the temperature according to RMM3. RMM3 writes some crap during overload, so Intel engineers asked to analyze it in EFI ... I
forgot to add, this is one of the latest top models.
Very disappointed with Intel after this...