FrForget

Thank you so much for replying. I’ve tried and it does indeed work. I was able to retrieve the ID using the unsigned device plugin too. Is using a vdisk slower than binding the IOMMU group? If I want to get the optimal performance, is this how I would achieve it? Thank you.

Hi everyone, I'm making some update to the way my Windows 11 VM works and it seems like I can't get it to work properly. Previously, I was using a 1 TB NVME drive where I was binding the corresponding VFIO at boot of the corresponding IOMMU group. This worked really well and I was able to select "None" as my Primary vDisk Location. Now, I decided to change my setup and little bit and want to use a SATA SSD as my main drive. I don't really care about my VM config so just decided to create a new one. I followed the same step, where I bind the corresponding VFIO at boot of the corresponding IOMMU group containing my SSD Drive. I have an MSI PRO B660-A DDR4 where all SATA ports are bound to 1 IOMMU group, which is fine, because I'm running my unraid array out of a separate LSI HBA card, so no issues there. In fact, it's even better because I'd love to pass not only a SATA SSD but some other SATA drives to my VM, so that would work. I start the VM, jump right into the Windows 11 installer, I can see my drive just fine, go along with the Windows 11 installation until it succeeds. Then, the VM restarts into the UEFI Interactive Shell, but I just can't see my drive. There is a few options, unfortunately, none seems to be corresponding to my SSD. At first, I thought maybe the binding was wrong, but then, the more I think about it, the more I was able to install Windows 11 on it, so the VM "should" be able to see it too right? From the UEFI Interactive Shell, I can type exit and then go to the BIOS Boot Manager, where I can see a few options. Windows Boot Manager : Flickers briefly and brings me back to the same BIOS Boot Manager Menu. EFI Internal Shell : Brings me back to the UEFI Interactive Shell. UEFI QEMU DVD-ROM QM00001 : Boots the Windows 11 installation. UEFI QEMU DVD-ROM QM00003 : Flickers briefly and brings me back to the same BIOS Boot Manager Menu. UEFI PXEv4 : Tries to boot from a network attached devices. I've tried playing around with the Logical CPUs, the Memory and a few other things. Here's what I currently have : CPU Mode : Host Passthrough Machine : Q35-7.1 BIOS : OVMF TPM Enable USB Boot : No Hyper-V : Yes USB Controller : 3.0 (qemu XHCI) Other PCI Devices : Intel Alder Lake-S PCH SATA Controller [AHCI Mode] | SATA controller (00:17.0) Is there anything else in my IOMMU group that I should pass to the VM for it to be able to boot from my drive? What am I missing? Please see the attached screenshots for more detail. Thank you!

Again, thank you so much! The information you've provided here are really useful.

Looks like it is completed as I can stop the array and see the full 1 TB now. I want to say a HUGE thank you for the instructions you've provided. Overall, would you say that this is the proper way to do it? If I wouldn't have change the slots number to 1 and simply unassigned the second drive and wait for it to balance, I wouldn't have had to use the command line do fix the drive right?

Interesting, now it says 1 TB! Now, previously, it was saying that a second drive was missing, does it now understand that only a single drive will remain in the cache?

Well, I think we are up to something here : Interestingly it's showing 564 GB as the size? Added new diags as well. crousti-server-diagnostics-20240412-1353.zip

Sorry, I can't start the array until I select the "Yes, I want to do this" saying Start will remove the missing cache disk and then bring the array on-line. That's what I want right ?

🙂

Alright, done! What's next 🙂? Going back to Main, I still see Unmountable. Do I need to Stop / Start the array? Reboot?

This is what I see now. Is it where I need to enter "write" ?

Here's the output : I haven't selected Yes or No. Too afraid to remove the data on the drive 🙃

There we go : Seems like 2 are found, is that right? However the sizes doesn't match.

Thank you for your answer. I just tried that and added the 1 TB back as the main cache pool drive. Unfortunately, I'm still seeing the same error : Then at the bottom, I also see this : Unmountable disk present: Cache • Samsung_SSD_980_PRO_1TB_S5P2NL0W409998F (nvme0n1) New diagnostics included too! crousti-server-diagnostics-20240412-1228.zip

Hi 🙂, I'm fairly new to unraid and everything seemed to be working well, until now. I decided that I wanted to migrate my cache pool to a new drive, going from 128 GB SSD to 1 TB NVME. I've seen online that a simple way to do it would be to add a new drive to the cache pool, wait for BTRFS sync to complete, stop the array, remove the 128 GB and voila, the 1 TB should now be the default drive in the cache pool. While the first few steps worked fine, I was indeed able to add the new drive to the cache pool, BTRFS synced happened, my array was then resized to 5XX GB (with 98% usage which seems suspicious), but I was able to stop the array, so I did, removed all drive from the cache pool, changed the slots to 1, added the 1 TB as the primary drive and started the array. It started, however I quickly saw this message next to the drive : Cache drive unmountable: Unsupported or no file system What are my options now? I haven't erased any drives and I've been trying to set the 128 GB back, or the 1TB, but haven't been successful. I included the diagnostics too! Thanks! crousti-server-diagnostics-20240412-1144.zip