pancho

Hello! Unable to passthrough Samsung 990 Evo plus NVME to a VM. Attempted all troubleshooting steps found in this forum, with no success. Summary below: Hardware Unraid 7.1.4 / Asrock H670m-ITX/AC---->i5-12400/32GB RAM 970 EVO (CACHE Drive - pcix3) / 990 EVO PLus NVME (For VM - pcix4) KVM IOMMU Groups - Validated NVME Controllers in separate Groups (990 EVO - Group 14) Q35-9.2 / OVMF-TPM VFIO - Bound 990 EVO Plus IOMMU group (attached) Primary Vdisk Location - NONE Other PCI Devices - Selected 990 EVO Plus (Boot Order '1') NVME Passthrough Problems (XML Attached) VM starts (no errors in logs). Hangs @ 'Unraid' splash screen. Single CPU core pegged @100% -- SLOOOOW - Takes 3-4 minutes before booting to Install ISO. No NVME Drives Found Troubleshooting Steps ISOLATED CPU CORES to VM - No change Swapped NVME Slots - No Change VFIO - Disabling bind to 990 SSD Controller SSD works without issue (VFIO) via Manual Vdisk - /dev/disk/by-id/nvme-Samsung_SSD_990_EVO_Plus Test - Outside of Unraid - Installed Windows 11 directly to 990 EVO PLUS - No issues Baffled why NVME passthrough 'seems to work', but hangs the VM & does not see 990 EVO Plus drive (i/o hardware limitation?). Can still use the VFIO method if needed, trying for best performance via passthrough. Suggestions? Thanks. Pancho VM.XML VM Log.txt

Upgraded today from rolledback version (6.8.3) to the latest version 6.9.2, unfortunately same issue as above where a single CPU core locks at @ 100% endlessly. The 100% Single CPU core occurred whether the array was loaded or not. Attached new diagnostics. Anyone have any suggestions? Thanks. mediabackup-diagnostics-20210408-1422.zip

Any Suggestions here? Thanks.

Apologies... re-uploaded. mediabackup-diagnostics-20210316-1208.zip

Hello! I currently have a backup Unraid server that I recently upgraded to version 6.9.1. Backup server spec's: ASRockRack C236 WSI (bios v2.70) i3-6100 CPU 8GB ECC RAM 2 VM's (all powered off) Docker - Not installed Post upgrade, the Dynamix S3 sleep function no longer worked (system would start sleep - and then immediately power back up). Further troubleshooting showed a single CPU core was maxed out @ 100%. The 100% Single CPU core occurred whether the array was loaded or not. Ran a diagnostics (attached). Rolled back to version 6.8.3, the CPU cores now operates properly (as does my sleep function). What is strange is that I have 2 other unraid servers (using identical motherboard) that work with S3 sleep (no CPU maxing out either). Is it possible this is an issue with the Flash boot drive (need to wipe clean and reload)? Any other Ideas? Thanks. Frank mediabackup-diagnostics-20210316-1208.zip

Hello! I am running ControlIR 2.9.1 and upgraded (from ver 6.4.0) to Unraid 6.5.0 today on one of my Unraid servers and I am no longer to see the running Dockers or VM's on my android app (S8+ - running Oreo),I just see the message 'No dockers are installed or they are currently unavailable. If I connect to another server (running version 6.4.0), I can see the Dockers/VM's without issue. Any suggestions on how to resolve? Thanks. Pancho