Encountered a issue reboot when upgrading pve version 8 to 9.

[phumstory]

PowerEdge T440 Tower Server - Log Error​

When rebooting it will hang, check the log error:

Category: System Health​

Severity: Critical
Message Id: PCI1318
Description: a fatal error was detected on a component at bus 4 device 0 function 1
(Log PCI1318 ที่เจอ (Bus 4, Dev 0, Func 1) → ตรงกับ PERC H730P ใน Slot 4)

**Version 8 does not encounter this issue.**

 

[Kingneutron]

T440 was introduced in 2017, PVE v9 kernel may have dropped support for the PERC card.

https://search.brave.com/search?q=d...summary=1&conversation=6ec7f411cd70ed85cc6a3d

You may want to replace the card with a newer model, check for firmware upgrades, or look for an updated version of perccli64

Also have the option to roll back to / restore PVE 8, it is supported until next year - and will give you some time to test upgrade to V9 on different hardware. You might be due for a hardware refresh as well, enterprise usually does it ~every 5 years or when support ends

 

[sightkick]

I have the same issue. Exchanged with a second "same" controller for troubleshooting purpose. Changes to HBA Mode with only one harddrive. Proxmox 9 fresh installed. But no success. From the first reboot on the error occurs.
I think the error appears since nearly one year!? But first on later Proxmox 8 versions.

Does anybody got a fix for it?
Which controller seems to be a good hardware refresh?

 

[sightkick]

The solution for me is found! It wasn´t the controller. The problem ist the Kerner >6.5 with the Broadcom network interface cards. The are not compatible. If you disable them in system Bios the machine ist booting fine (PVE9) without any error AND with the Perc H730p working. Check out the reddit thread: https://www.reddit.com/r/Proxmox/comments/1o6lrh2/server_not_rebooting_properly/

Now I put in another network card and everything is fine!

Hope this helps every DELL T440 User to give their server a new job