The server, as scheduled at 10 Moscow time and 22 Moscow time, crashes with LA>50. Where to dig?

Question

linux

No ideas, no traffic during these periods, crontab looked. The rest of the time it works perfect. I just don't understand what's going on, can someone tell me what?

Answer 1

50+ LA is the average temperature in the hospital. I have one of the carcasses under combat load worth 1300 la and joyfully answers everyone.
But if your wa is more than 5-10%, dig in the direction of slowing down disks or networks. Well, another IO, if it is there.

Answer 2

/var/spool/cron/crontabs/* also look if you are sinning on cron.

Answer 3

And what's in the logs? Clean reboot or something really crashed?
If it's still not clear, then you can try to put all the processes manually right before the X hour and see if the clean system reboots by itself.
And it looks like some kind of maintenance script.

Answer 4

With these 50+, the web server does not respond at all and the console even slows down. This is how many processors you have, then 1300 is ok?

Answer 5

I'm guessing it has something to do with backup. Purely theoretical.
Start with /etc/cron.dor where do you have crons?

Answer 6

Did you look at the status and logs of the raid controller? Mb it at this time verify does. I suspect that reducing the responsiveness of the disks will calm it down.