Random crashes + Can't boot without a GPU and monitor (6.12.4)

[JamesyBoy]

Hey all. I've been playing with Unraid for about 2 weeks now, and have been trying to troubleshoot 2 frustrating problems. Thanks for any help you can offer!

 

1 - My server won't boot unless I move over my GTX 2080 from my gaming computer and attach a monitor. Once attached, it boots from the USB and starts without needing any input from my end. I've been able to subsequently remove the GPU after it boots. But if I boot the server without the GPU, the server doesn't start. Not really sure what's happening there. My CPU doesn't have integrated graphics, so I can't attach a monitor that way.

 

2 - Upon booting, my server will last anywhere from 2 to 36 hours before crashing. I've been right beside it when it's happened several times; the fans and lights are all still on, except it can't be reached, and it's gone from my router's Connected Devices panel. When I turn it off and on (after reconnecting my GPU and monitor), it boots fine and then needs a Parity check, which is starting to get tedious.

 

Steps tried to no avail:

• The server has 2 x 32GB RAM. I've tried running with both sticks, as well as each stick individually.
• Updated the motherboard BIOS
• Copying logs to the USB stick. After the most recent crash, there were no entries within 10min of the crash, so I'm not sure if there's much useful. I can post the full log if it would be helpful.

 

Next steps:

• reseating the CPU?
• replacing the CPU with one with integrated graphics?
• replacing the CPU and mobo with newer models?
• any other ideas?

 

Thanks!

 

System Specs:

M/B:    Micro-Star International Co., Ltd. B350M PRO-VDH (MS-7A38) Version 2.0 s/n H416735843
BIOS:    American Megatrends International, LLC. Version A.L4 Dated 05/17/2023
CPU:    AMD Ryzen 5 1600 Six-Core @ 3200 MHz (no integrated graphics)

Drives:    2 x 8TB Seagate drives
HVM:    Enabled
IOMMU:    Enabled
Cache:    L1 - Cache: 576 KiB, L2 - Cache: 3 MiB, L3 - Cache: 16 MiB
Memory:    32 GiB DDR4 (max. installable capacity 128 GiB)
Network:    bond0: fault-tolerance (active-backup), mtu 1500
Kernel:    Linux 6.1.49-Unraid x86_64
OpenSSL:    1.1.1v

[trurl]

 

[JamesyBoy]

Thanks for the link trurl! I will try disabling C-States and see what happens.

 

24 hour update: still running!

4 day update: still running!

 

(if anyone has any suggestions about my first GPU required problem, let me know)

Edited October 15, 2023 by JamesyBoy

[ConnerVT]

Check your BIOS settings for GPU/Display/Multi Monitor/ etc, and try different choices for one that hopefully works.

 

If no success with BIOS settings, toss any cheap GPU in the machine to make it happy.  I have a Zoltac GT710 in my backup server (Ryzen 1500X) that's only there for troubleshooting.  If you need buy a cheap card, a Quadro P400 can be had for less than $100 USD new, half that used.

 

(Fun Fact: The GT 710 was the lowest capable card that Microsoft allowed one to install Windows 10.  That's why I had one on hand.)