Im having a strange issue. When i upgrade to Unraid 7 from Unraid 6, none of my array drives connected to my LSI HBA are detected (both nvme cache drives are working). Array is stopped on boot and start array says there are missing disks. All drives parity+data show unassigned with no drives available in the drop-down.

 

 

My Setup is:

Motherboard: Asus Pro W680 + IPMI

CPU: Intel i9-14900k

Array connected via two LSI LSI 9207-8i cards in in IT Mode

Parity: 

24TB Seagate

Data drives:

3x 8TB WD data drives

2x 24TB Seagate

 

This is the second time Ive attempted this upgrade. Last time I had the same issue and just rolled back and everything worked again. I'm sure is something with Unraid 7. 

 

My HBA cards are both visible in the ASUS bios on Legacy BIOS config.

 

Another issue, perhaps related, is my server wont power down. It sends the commands and terminal says SHUTDOWN, but it just stalls. I can figure this out later after I get my server in a better state, but wanted to share here in case its tickles someones brain and is related.

 

Anyone have any ideas? Ive searched around and came up blank. Ive also read the upgrade forum and none of those known issues sould like what i'm expiriencing.

 

 

Edited April 20, 20251 yr by BrickPotato
formatting and corrected prior version from 6.9.12 to 6.12.4 in title

Here is what the array looks like after boot:

Also seeing this error spam the logs every second. These are the devices connected to the HBA:

Apr 20 13:22:08 Potato-Server emhttpd: device /dev/sdh problem getting id
Apr 20 13:22:08 Potato-Server emhttpd: device /dev/sdg problem getting id
Apr 20 13:22:08 Potato-Server emhttpd: device /dev/sdd problem getting id
Apr 20 13:22:08 Potato-Server emhttpd: device /dev/sde problem getting id
Apr 20 13:22:08 Potato-Server emhttpd: device /dev/sdf problem getting id
Apr 20 13:22:08 Potato-Server emhttpd: device /dev/sdc problem getting id

Apr 20 13:22:09 Potato-Server emhttpd: device /dev/sdh problem getting id
Apr 20 13:22:09 Potato-Server emhttpd: device /dev/sdg problem getting id
Apr 20 13:22:09 Potato-Server emhttpd: device /dev/sdd problem getting id
Apr 20 13:22:09 Potato-Server emhttpd: device /dev/sde problem getting id
Apr 20 13:22:09 Potato-Server emhttpd: device /dev/sdf problem getting id
Apr 20 13:22:09 Potato-Server emhttpd: device /dev/sdc problem getting id

 

 

Here is my config prior to the upgrade:

 

I ran dmesg | grep -i sas and it looks like both of my LSI 9207-8i HBAs are being detected. The kernel loads the mpt3sas driver successfully and both controllers initialize with port enable: SUCCESS. It logs all 6 attached drives under SCSI with valid SAS addresses, and I see this for each:

 

scsi 3:0:X:0: SATA: handle(...), sas_addr(...), phy(X), device_name(0x0000000000000000)

 

Despite this, Unraid is still spamming the log every second with:

 

emhttpd: device /dev/sdX problem getting id

 

for all 6 drives. These same drives show up correctly in Unraid 6.12.4 with no issues, so this (to me) looks like a Unraid 7 issue with how it identifies drives from the HBA.

Let me know if there's anything else I can try. Happy to pull more logs or test patches.

 

Found this post with almost my exact same config (motherboard, HBAs, etc):

 

Going to troubleshoot based on their findings. Will report back.

Attach Diagnostics ZIP to your NEXT post in this thread.x

Turns out the issue wasn’t Unraid or the HBAs directly... it was CSM being disabled in my motherboard BIOS. (Doh!)

 

After enabling CSM, Unraid recognized all my drives connected to the LSI 9207-8i cards. The “problem getting id” errors disappeared, and the array started up with no issues.

 

I've attached the diagnostics file I was looking over while the issue was still occurring. Strange that this is a change in behavior between Unraid 6 < - > Unraid 7, but happy to have it figured out. My motherboard doesn't let me use iGPU wish CSM enabled, so that's something else I'll have to figure out.

diagnostics-20250420-1333.zip