November 24, 20241 yr Hi! I have this issue in my array and I have tried every possible fix that I could find on the site. I have version 6.12.13. Here is the list of things that I have already tried: 1) Shutdown and boot again 2) Replace the SATA cable and connect to a different SATA port on the motherboard 3) Parity check 4) Stop the array, leave the disk unassigned, start the array; then stop, assign the disk to the array and let it rebuild from parity 5) Unassign the disk again, format it, ple-clear, assign it to the array and let it rebuild from parity. On this ocassion it seemed to be fine at first, but about 15 minutes later it showed again "disk is disabled, contents emulated" The ple-clear report in step 5 showed 0 errors. SMART report shows no errors. What else can I try if everything tells me the disk is fine??
November 24, 20241 yr Try having it rebuild the disk again from parity, and if/when it fails and becomes disabled during the process, post your diagnostics
November 24, 20241 yr Author Thanks, Squid. I will try that. At the moment, as a safety precaution, I'm using unbalance to scatter the files from that disk to the rest of the array (I have enough space). Once the process is finished, I will remove the disk from the array and add it again as if it was a new one... Then see what happens and share the diagnostics here. Fingers crossed.
November 24, 20241 yr Community Expert 3 hours ago, impoftheperverse said: 3) Parity check Can't really do a "parity check" when you only have single parity and a disabled disk. It will only do a read check. 3 hours ago, impoftheperverse said: 5) Unassign the disk again, format it, ple-clear Pointless to format then preclear, and no reason to do either before rebuild. At least you didn't try to format the emulated disk.
November 26, 20241 yr Author OK... I did the parity rebuild again, 0 errors again, all green, all fine for a few minutes. Then I started the VM (windows 10, with the disk running from an unassigned NVDME drive) and suddenly read errors started to pile up, this time in the parity disk! The disk that was wrong/disabled previously now is fine, so I have no idea about what's going on, but definitely has nothing to do with the physical drives. I could not capture the diagnostics at the time of the failure, as both the VM went unusable and the server page freeze no matter what I tried. I had to forze shutdown and boot again, but here attached are the diagnostics from the fresh server start. impvault-diagnostics-20241126-1930.zip
November 26, 20241 yr Community Expert Solution 09:00.0 SATA controller [0106]: Advanced Micro Devices, Inc. [AMD] FCH SATA Controller [AHCI mode] [1022:7901] (rev 51) Subsystem: Advanced Micro Devices, Inc. [AMD] FCH SATA Controller [AHCI mode] [1022:7901] Kernel driver in use: ahci Kernel modules: ahci This controller is being passed through to a VM, so when the VM starts Unraid will lose connection with all connected disks, just correct that.
November 26, 20241 yr Author Thanks a lot, Jorge, for having a look at it. Could you point me in the right direction as to how to prevent that controller from being passed through to the VM? I follow lots of tutorials, but my understanding of how things works is pretty shallow... Sorry. (In case it's important, I'm using PCIe ACS Override in system devices, as that is the only way I can pass through my GPU) PS. once that is solved, should I stop/start the array and perform a parity check or how do I get parity enabled again? Thanks! Edited November 26, 20241 yr by impoftheperverse
November 27, 20241 yr Community Expert Remove this section from the VM XML <hostdev mode="subsystem" type="pci" managed="yes"> <driver name="vfio"/> <source> <address domain="0x0000" bus="0x09" slot="0x00" function="0x0"/> </source> <address type="pci" domain="0x0000" bus="0x00" slot="0x09" function="0x0"/> </hostdev>
November 27, 20241 yr Author I removed that code from the VM which was causing the issue, and I also checked the other 2. However, when I rebooted, without even starting the VMs, the vfio-pci log shows that the controller is still being binded. I attached full diagnostics. Btw. when I rebooted a warning poped up saying "VM Autostart disabled due to vfio-bind error". I don't have any of my VMs in autostart... Loading config from /boot/config/vfio-pci.cfg BIND=0000:09:00.0|10de:1e84 0000:09:00.1|10de:10f8 0000:09:00.2|10de:1ad8 0000:09:00.3|10de:1ad9--- Processing 0000:09:00.0 10de:1e84 --- Processing 0000:09:00.1 10de:10f8 --- Processing 0000:09:00.2 10de:1ad8 --- Processing 0000:09:00.3 10de:1ad9 --- Devices listed in /sys/bus/pci/drivers/vfio-pci: vfio-pci binding complete impvault-diagnostics-20241127-1900.zip Edited November 27, 20241 yr by impoftheperverse
November 28, 20241 yr Author On 11/27/2024 at 7:25 PM, JorgeB said: It's also on /config/vfio-pci.cfg, deleted that file Thank you once again. Everything was solved. Not sure if it was related, but before this happened I physically removed an old GPU and 2 (unassigned) ssd drives from the server case. I suppose the problem with the SATA controller being passed through had somehow something to do with that. Edited November 28, 20241 yr by impoftheperverse
November 29, 20241 yr Community Expert When you add or remove hardware, the devices IDs will typically change, no you need to recheck all bindings/devices being passed through.
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.