As mentioned above, I got around the CPU lockup problem by installing CentOS kernel-5.14.0-302.el9. That has been running well for at least a couple weeks.
I did a dnf check-update just now and it shows an update for kernel-devel-matched 5.14.0-284.18.1.el9_2.
Anybody know if I do the dnf update is it going to downgrade my system to kernel 5.14.0-284.18.1.el9_2? 284 is the kernel that causes the CPU lockup problem so definitely don’t want a downgrade!
I just noticed that the dnf provided kernel is actually newer: 284.18.1 versus 284.11.1. Anybody know if 284.18.1 fixes the CPU lockup bug?
1 Like
@hspindel I’m also wondering if version 284.18.1 solves this problem. I’m currently running my machines on 162.23.1. 284.11.1 caused all three, which all had different CPU architectures (Opteron, Epyc, and Xeon) to lock repeatedly. I’m a little gun-shy of installing another 284…
I had to reinstall my Rocky Linux server from scratch, so I gave 284.18.1 a try (as above, had been using a CentOS kernel). Server has been up for over 24 hours without an issue, so my guess is that 284.18.1 did fix the issue seen with 284.11.1.
3 Likes
I created an account here to remind everyone to remember to disable Fast Boot in the Bios as one of your diagnotic steps. Without that, ie. normal boot up, the system will check that each peice of hardware connected to the motherboard is properly repsonding before you start up. That is not a bad thing. And with SSD’s today, the difference is typically only seconds anyway, to potentially avoid unecessay greif of going down the rabbit hole of which piece of hareware is causing this reboot… I hope this helps many users who see this. It stopped my randome reboots on Rocky Linux 9.3 when I changed from a Fedora 39 host ot Rocky Linux 9.3.