VM Win10 Weirdness

[Julex]

I'm running a Windows 10 VM on Unraid 6.6.0, Primary drive is passed-through.  I'm having a couple hiccups, that didn't happen in a previously setup VM. 

 

The first issue, after leaving VM running all night, when I come back to it in the morning, there is no response.  Screen stays idle, tried tapping keys and mouse strokes, no response.  I am unable to connect with splashtop as it says the VM is off.  When I go into Unraid/VMs, it's listed as running.  Trying to Hibernate comes up with "

Domain 'x' could not be suspended. Guest agent is not responding: QEMU guest agent is not connected.", Pausing and Starting, Stop, and Restart appear to do nothing.  Force Stop is my only alternative.  The question now is, what could be going on?

 

The second one, even with the keyboard check mark passed through from the GUI, when the VM starts, the keyboard does not work.  I have to go into the Hotplug USB plugin and attach keyboard. 

[Julex]

Welp, this may explain some more, looking at unassigned devices, my Primary VM drive shows used and free space in MBs, rather than GBs...   the log shows 

Dec 16 11:22:48 DarkPrism kernel: ata1: SATA max UDMA/133 abar m2048@0xe6136000 port 0xe6136100 irq 27
Dec 16 11:22:48 DarkPrism kernel: ata1: SATA link up 6.0 Gbps (SStatus 133 SControl 300)
Dec 16 11:22:48 DarkPrism kernel: ata1.00: supports DRM functions and may not be fully accessible
Dec 16 11:22:48 DarkPrism kernel: ata1.00: disabling queued TRIM support
Dec 16 11:22:48 DarkPrism kernel: ata1.00: ATA-9: Samsung SSD 840 EVO 250GB, S1DBNSAF835247V, EXT0DB6Q, max UDMA/133
Dec 16 11:22:48 DarkPrism kernel: ata1.00: 488397168 sectors, multi 1: LBA48 NCQ (depth 32), AA
Dec 16 11:22:48 DarkPrism kernel: ata1.00: supports DRM functions and may not be fully accessible
Dec 16 11:22:48 DarkPrism kernel: ata1.00: disabling queued TRIM support
Dec 16 11:22:48 DarkPrism kernel: ata1.00: configured for UDMA/133
Dec 16 11:22:48 DarkPrism kernel: sd 1:0:0:0: [sdh] 488397168 512-byte logical blocks: (250 GB/233 GiB)
Dec 16 11:22:48 DarkPrism kernel: sd 1:0:0:0: [sdh] Write Protect is off
Dec 16 11:22:48 DarkPrism kernel: sd 1:0:0:0: [sdh] Mode Sense: 00 3a 00 00
Dec 16 11:22:48 DarkPrism kernel: sd 1:0:0:0: [sdh] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Dec 16 11:22:48 DarkPrism kernel: sdh: sdh1 sdh2 sdh3 sdh4
Dec 16 11:22:48 DarkPrism kernel: sd 1:0:0:0: [sdh] Attached SCSI disk
Dec 16 11:23:13 DarkPrism emhttpd: Samsung_SSD_840_EVO_250GB_S1DBNSAF835247V (sdh) 512 488397168
Dec 16 11:23:30 DarkPrism unassigned.devices: Adding disk '/dev/sdh1'...
Dec 16 11:23:30 DarkPrism unassigned.devices: Mount drive command: /sbin/mount -t ntfs -o auto,async,noatime,nodiratime,nodev,nosuid,umask=000 '/dev/sdh1' '/mnt/disks/Recovery'
Dec 16 11:23:30 DarkPrism unassigned.devices: Mount drive ro command: /sbin/mount -t ntfs -ro auto,async,noatime,nodiratime,nodev,nosuid,umask=000 '/dev/sdh1' '/mnt/disks/Recovery'
Dec 16 11:23:30 DarkPrism unassigned.devices: Adding disk '/dev/sdh2'...
Dec 16 11:23:30 DarkPrism unassigned.devices: Mount drive command: /sbin/mount -t vfat -o auto,async,noatime,nodiratime,nodev,nosuid,umask=000 '/dev/sdh2' '/mnt/disks/Samsung_SSD_840_EVO_250GB_S1DBNSAF835247V-part2'
Dec 16 11:23:30 DarkPrism unassigned.devices: Successfully mounted '/dev/sdh2' on '/mnt/disks/Samsung_SSD_840_EVO_250GB_S1DBNSAF835247V-part2'.
Dec 16 11:23:30 DarkPrism unassigned.devices: Device '/dev/sdh2' script file not found. 'ADD' script not executed.
Dec 16 11:23:31 DarkPrism unassigned.devices: Adding disk '/dev/sdh3'...
Dec 16 11:23:31 DarkPrism unassigned.devices: No filesystem detected on '/dev/sdh3'.
Dec 16 11:23:31 DarkPrism unassigned.devices: Adding disk '/dev/sdh4'...
Dec 16 11:23:31 DarkPrism unassigned.devices: Mount drive command: /sbin/mount -t ntfs -o auto,async,noatime,nodiratime,nodev,nosuid,umask=000 '/dev/sdh4' '/mnt/disks/Samsung_SSD_840_EVO_250GB_S1DBNSAF835247V-part4'
Dec 16 11:23:31 DarkPrism unassigned.devices: Mount failed with error: ntfs_mst_post_read_fixup_warn: magic: 0x00000000 size: 1024 usa_ofs: 0 usa_count: 0: Invalid argument Record 10 has no FILE magic (0x0) Failed to open inode FILE_UpCase: Input/output error Failed to mount '/dev/sdh4': Input/output error NTFS is either inconsistent, or there is a hardware fault, or it's a SoftRAID/FakeRAID hardware. In the first case run chkdsk /f on Windows then reboot into Windows twice. The usage of the /f parameter is very important! If the device is a SoftRAID/FakeRAID then first activate it and mount a different device under the /dev/mapper/ directory, (e.g. /dev/mapper/nvidia_eahaabcc1). Please see the 'dmraid' documentation for more details.
Dec 16 11:23:31 DarkPrism unassigned.devices: Mount drive ro command: /sbin/mount -t ntfs -ro auto,async,noatime,nodiratime,nodev,nosuid,umask=000 '/dev/sdh4' '/mnt/disks/Samsung_SSD_840_EVO_250GB_S1DBNSAF835247V-part4'
Dec 16 11:23:31 DarkPrism unassigned.devices: Mount of '/dev/sdh4' failed. Error message: ntfs_mst_post_read_fixup_warn: magic: 0x00000000 size: 1024 usa_ofs: 0 usa_count: 0: Invalid argument Record 10 has no FILE magic (0x0) Failed to open inode FILE_UpCase: Input/output error Failed to mount '/dev/sdh4': Input/output error NTFS is either inconsistent, or there is a hardware fault, or it's a SoftRAID/FakeRAID hardware. In the first case run chkdsk /f on Windows then reboot into Windows twice. The usage of the /f parameter is very important! If the device is a SoftRAID/FakeRAID then first activate it and mount a different device under the /dev/mapper/ directory, (e.g. /dev/mapper/nvidia_eahaabcc1). Please see the 'dmraid' documentation for more details.
Dec 16 11:25:26 DarkPrism unassigned.devices: Device '/dev/sdh2' script file not found. 'REMOVE' script not executed.
Dec 16 11:25:26 DarkPrism unassigned.devices: Unmounting '/dev/sdh2'...
Dec 16 11:25:26 DarkPrism unassigned.devices: Unmount cmd: /bin/umount '/dev/sdh2' 2>&1
Dec 16 11:25:26 DarkPrism unassigned.devices: Successfully unmounted '/dev/sdh2'
Dec 16 11:25:30 DarkPrism unassigned.devices: Adding disk '/dev/sdh1'...
Dec 16 11:25:30 DarkPrism unassigned.devices: Mount drive command: /sbin/mount -t ntfs -o auto,async,noatime,nodiratime,nodev,nosuid,umask=000 '/dev/sdh1' '/mnt/disks/Recovery'
Dec 16 11:25:30 DarkPrism unassigned.devices: Mount drive ro command: /sbin/mount -t ntfs -ro auto,async,noatime,nodiratime,nodev,nosuid,umask=000 '/dev/sdh1' '/mnt/disks/Recovery'
Dec 16 11:25:30 DarkPrism unassigned.devices: Adding disk '/dev/sdh2'...
Dec 16 11:25:30 DarkPrism unassigned.devices: Mount drive command: /sbin/mount -t vfat -o auto,async,noatime,nodiratime,nodev,nosuid,umask=000 '/dev/sdh2' '/mnt/disks/Samsung_SSD_840_EVO_250GB_S1DBNSAF835247V-part2'
Dec 16 11:25:30 DarkPrism unassigned.devices: Successfully mounted '/dev/sdh2' on '/mnt/disks/Samsung_SSD_840_EVO_250GB_S1DBNSAF835247V-part2'.
Dec 16 11:25:30 DarkPrism unassigned.devices: Device '/dev/sdh2' script file not found. 'ADD' script not executed.
Dec 16 11:25:30 DarkPrism unassigned.devices: Adding disk '/dev/sdh3'...
Dec 16 11:25:30 DarkPrism unassigned.devices: No filesystem detected on '/dev/sdh3'.
Dec 16 11:25:30 DarkPrism unassigned.devices: Adding disk '/dev/sdh4'...
Dec 16 11:25:30 DarkPrism unassigned.devices: Mount drive command: /sbin/mount -t ntfs -o auto,async,noatime,nodiratime,nodev,nosuid,umask=000 '/dev/sdh4' '/mnt/disks/Samsung_SSD_840_EVO_250GB_S1DBNSAF835247V-part4'
Dec 16 11:25:30 DarkPrism unassigned.devices: Mount failed with error: ntfs_mst_post_read_fixup_warn: magic: 0x00000000 size: 1024 usa_ofs: 0 usa_count: 0: Invalid argument Record 10 has no FILE magic (0x0) Failed to open inode FILE_UpCase: Input/output error Failed to mount '/dev/sdh4': Input/output error NTFS is either inconsistent, or there is a hardware fault, or it's a SoftRAID/FakeRAID hardware. In the first case run chkdsk /f on Windows then reboot into Windows twice. The usage of the /f parameter is very important! If the device is a SoftRAID/FakeRAID then first activate it and mount a different device under the /dev/mapper/ directory, (e.g. /dev/mapper/nvidia_eahaabcc1). Please see the 'dmraid' documentation for more details.
Dec 16 11:25:30 DarkPrism unassigned.devices: Mount drive ro command: /sbin/mount -t ntfs -ro auto,async,noatime,nodiratime,nodev,nosuid,umask=000 '/dev/sdh4' '/mnt/disks/Samsung_SSD_840_EVO_250GB_S1DBNSAF835247V-part4'
Dec 16 11:25:30 DarkPrism unassigned.devices: Mount of '/dev/sdh4' failed. Error message: ntfs_mst_post_read_fixup_warn: magic: 0x00000000 size: 1024 usa_ofs: 0 usa_count: 0: Invalid argument Record 10 has no FILE magic (0x0) Failed to open inode FILE_UpCase: Input/output error Failed to mount '/dev/sdh4': Input/output error NTFS is either inconsistent, or there is a hardware fault, or it's a SoftRAID/FakeRAID hardware. In the first case run chkdsk /f on Windows then reboot into Windows twice. The usage of the /f parameter is very important! If the device is a SoftRAID/FakeRAID then first activate it and mount a different device under the /dev/mapper/ directory, (e.g. /dev/mapper/nvidia_eahaabcc1). Please see the 'dmraid' documentation for more details.

[UUB86]

On 12/16/2018 at 11:35 AM, Julex said:

Dec 16 11:25:30 DarkPrism unassigned.devices: Mount of '/dev/sdh4' failed. Error message: ntfs_mst_post_read_fixup_warn: magic: 0x00000000 size: 1024 usa_ofs: 0 usa_count: 0: Invalid argument Record 10 has no FILE magic (0x0) Failed to open inode FILE_UpCase: Input/output error Failed to mount '/dev/sdh4': Input/output error NTFS is either inconsistent, or there is a hardware fault, or it's a SoftRAID/FakeRAID hardware. In the first case run chkdsk /f on Windows then reboot into Windows twice. The usage of the /f parameter is very important! If the device is a SoftRAID/FakeRAID then first activate it and mount a different device under the /dev/mapper/ directory, (e.g. /dev/mapper/nvidia_eahaabcc1). Please see the 'dmraid' documentation for more details.

 

I'm also running into the exact same problem with unraid 6.6.6.  I'm passing through an ssd directly to the VM(I'm also passing through a GPU if that matters).  For me I set up a brand new VM and am using OVMF as my bios for the VM.  I can restart the VM however once I shut down the VM, I won't be able to start it again because of the above error.  Before shutting down, the VM works fine and there is no indication that anything is wrong.  With my previous VM I was doing roughly the same thing the only differences being seabios and a different gpu.  Could the bios be causing this issue?  I was using a radeon gpu but I upgraded to an nvidia gpu and I was hoping to not have to reboot unraid every time the VM needs a reboot.  I've been struggling with this since last week and this seems to be the first related forum post that I could find.  If anyone has any ideas I would appreciate it.  If needed I can provide more logs/vm details once I get home from work.

[Symon]

Were you guys able to solve this issue? Im currently getting simmilar errors on the log:

Jan 23 20:34:34 Tower ntfs-3g[7725]: ntfs_mst_post_read_fixup_warn: magic: 0x61636f6c size: 4096 usa_ofs: 8300 usa_count: 11333: Invalid argument
Jan 23 20:34:34 Tower ntfs-3g[7725]: Actual VCN (0x286b6361706e7520) of index buffer is different from expected VCN (0x0) in inode 0x389b.
Jan 23 20:34:45 Tower ntfs-3g[7725]: ntfs_mst_post_read_fixup_warn: magic: 0x54484346 size: 4096 usa_ofs: 46002 usa_count: 19978: Invalid argument
Jan 23 20:34:45 Tower ntfs-3g[7725]: Actual VCN (0x72657473696e6953) of index buffer is different from expected VCN (0x0) in inode 0x5cbb.
Jan 23 20:34:45 Tower ntfs-3g[7725]: ntfs_mst_post_read_fixup_warn: magic: 0x2c275527 size: 4096 usa_ofs: 12891 usa_count: 23857: Invalid argument
Jan 23 20:34:45 Tower ntfs-3g[7725]: Actual VCN (0x325b2c2757273d5d) of index buffer is different from expected VCN (0x1) in inode 0x3562.
Jan 23 20:34:45 Tower ntfs-3g[7725]: ntfs_mst_post_read_fixup_warn: magic: 0x22657275 size: 4096 usa_ofs: 8236 usa_count: 11565: Invalid argument
Jan 23 20:34:45 Tower ntfs-3g[7725]: Actual VCN (0x756e6f4274654722) of index buffer is different from expected VCN (0x0) in inode 0x364d.
Jan 23 20:34:45 Tower ntfs-3g[7725]: ntfs_mst_post_read_fixup_warn: magic: 0x5b5b2d2d size: 4096 usa_ofs: 11565 usa_count: 11565: Invalid argument
Jan 23 20:34:45 Tower ntfs-3g[7725]: Actual VCN (0x2d2d2d2d2d2d2d2d) of index buffer is different from expected VCN (0x0) in inode 0x4d72.
Jan 23 20:34:45 Tower ntfs-3g[7725]: ntfs_mst_post_read_fixup_warn: magic: 0x2069553c size: 4096 usa_ofs: 28024 usa_count: 28268: Invalid argument
Jan 23 20:34:45 Tower ntfs-3g[7725]: Actual VCN (0x6c622e7777772f2f) of index buffer is different from expected VCN (0x0) in inode 0x3840.

If I unmount/mount the disk the error messages dissapear but reapear after some time .. 

Running 6.6.6

[Julex]

MY issue was caused by mounting my unassigned disk and using as a pass-through device.  Leave drive unmounted if passing unassigned drive through to VM. 

 

Lack of knowledge on how unassigned devices work in unraid, led me to believe that drives must be mounted in order for unraid to see them.  This is not the case, and in fact causes massive problems, IE unraid failing to see drive partitions after reboot.

Edited January 30, 2019 by Julex