My Windows 11 Blue Iris VM has been running fine until it suddenly now keeps getting paused. It won't resume, but when I force shutdown and start it it will run long enough for me to RDP and everything looks OK in the client (no power savings turned on, plenty of disk space), but it's only lasting a matter of minutes before getting paused again.  In the libvirt diagnostic file I've just seen this:

 

2025-02-08 11:03:27.912+0000: 12703: warning : qemuProcessHandleIOError:866 : Transitioned guest vm-blueiris to paused state due to IO error
2025-02-09 11:37:07.454+0000: 12703: warning : qemuProcessHandleIOError:866 : Transitioned guest vm-blueiris to paused state due to IO error
2025-02-09 11:37:20.045+0000: 12703: warning : qemuProcessHandleIOError:866 : Transitioned guest vm-blueiris to paused state due to IO error

 

Is there any way to get more info on this, at least which disk is the issue? It's using the cache pool for its C drive, and that's 2x 1TB NVME SSDs with btrfs. They both show 0 errors.

 

The main video recording drive is a WD Purple 6TB HDD, that's accessed as an Automounted Unassigned Device. It shows full in Unraid but in the VM I can see it has plenty of free space. The dashboard isn't showing a disk error count, but when I run a SMART short test (extended greyed out) it shows 0 errors. 

 

EDIT - when I do a Tools -> Check Disk on the both drives from within the VM, the C drive scan finished fine, but when I do it on the video storage drive, it locks up the whole VM in seconds. So I guess I've answered that question. Seems weird though, that such an issue with a non-OS drive can completely lock up Windows/the VM, and also that Unraid itself doesn't see an issue with the drive (at least the quick SMART scan).

 

I'm on 7.0.0.

 

Would appreciate any advice as I'm without my security recording system until I get this fixed!

Edited February 9, 20251 yr by Richard Harnwell
new info

The vdisk is likely fully allocated, since it's thin provisioned, recommend passing through the actual disk and let the VM have access to it.

1 hour ago, JorgeB said:

The vdisk is likely fully allocated, since it's thin provisioned, recommend passing through the actual disk and let the VM have access to it.

Thanks. Not really got my head around the subtleties of allocating a disk to a VM though, so no quite sure what you mean. The disk was certainly showing 1TB of free space in the guest VM though.

 

I have since found another drive and allocated that and it seems OK so far. I think I've obviously allocated it slightly differently this time though, as this time it doesn't show the used percentage in Unraid. In the VM settings I'm referencing it as /dev/disk/by-id/ata.. Would you consider this the correct/best way?

 

 

Now the old/problematic disk (6TB) isn't in use for Blue Iris I started a SMART extended self test on it. It has seemingly locked up at 10% (no obvious way to interrupt it). So I guess maybe there is a fundamental issue with it?

Edited February 9, 20251 yr by Richard Harnwell

19 hours ago, Richard Harnwell said:

Would you consider this the correct/best way?

Yep.