Spies

Does unraid actually use reflink in any way?

No they're not encrypted, it does seem excessve to lose 14GB of each drive though!

Remove half the sticks, see if the error count changes. Maybe try with just the crucial memory.

I've set up a second server with 2*2TB drives in xfs format, I've disabled docker and VM and deleted the dcoker and libvirt images, but I'm still seeing 14GB used on both discs which are currently in the array (no parity for now). I understand xfs has an overhead of 1GB for TB but 14GB seems excessive. On browsing the drives, all I can see are the 4 folders it creates, but nothing inside them, where is the space being used?

Turns out, it was because I had set an MTU lower than 1500 under Settings -> Network Settings, once set back to 1500, everything works again!

I cannot load the web interfaces for either my openwrt router or synology, both of which use nginx to serve the pages, is there any conceivable reason why this might be the case? It happens in both my Windows 10 VM and MacOS VM, both using br0 for their network access. I've done some searching, but most hits referring to issues to hosting an nginx server on KVM, not trying to load one from another networked device! The Samba share on the synology hardly works from the VM also, whereas it works fine on another, baremetal machine without issue.

I have several hard drives which have been rotated out of my server for bigger ones, I don't have any more bays in my case, I don't have enough pcie slots either as I'm not using server grade hardware (and don't intend to upgrade any time soon). So, I would need to get either a Silverstone 380B or Fractal Node 304. This leaves me in a tricky position in the quest for more space, my Dell perc already has both connections used, I can't put another HBA in as I don't have any full size pcie slots left and even if I did, it would be getting filled with an Nvidia card for Plex transcoding. I could remove one of the connections on the perc, connect that to a floating expander in the other case and then run a cable back in for the 4 drives that were connected originally. Or I could build a second low powered machine, as it will only be a NAS and map the drives to my main server using unassigned devices. What would you do?

It works on the rear USB 3.0 ports.

Interestingly, if I use pre-clear in unraid on the same device, on the same port, there are no issues, anyone know whats going on? Edit: Scratch that, the syslog has filled up with Jun 6 17:01:40 Tower kernel: Buffer I/O error on dev sdm, logical block 122096640, async page read Jun 6 17:01:40 Tower kernel: Buffer I/O error on dev sdm, logical block 122096640, async page read Jun 6 17:01:40 Tower kernel: Buffer I/O error on dev sdm, logical block 122096640, async page read Jun 6 17:01:40 Tower kernel: Buffer I/O error on dev sdm, logical block 122096640, async page read Jun 6 17:01:40 Tower kernel: Buffer I/O error on dev sdm, logical block 122096640, async page read Jun 6 17:01:40 Tower kernel: Buffer I/O error on dev sdm, logical block 122096640, async page read Jun 6 17:01:40 Tower kernel: Buffer I/O error on dev sdm, logical block 122096640, async page read Jun 6 17:01:40 Tower kernel: Buffer I/O error on dev sdm, logical block 122096640, async page read Jun 6 17:01:40 Tower kernel: Buffer I/O error on dev sdm, logical block 122096640, async page read Jun 6 17:01:40 Tower kernel: Buffer I/O error on dev sdm, logical block 122096640, async page read Jun 6 17:01:41 Tower kernel: usb 2-4: device descriptor read/8, error -110

I need an additional full size PCIe slot, are both the aforementioned motherboards fully compatible with unraid? I use an NVME drive, so I understand that on the B450-F I'd have to use m2.2 and that would halve the bandwidth of the 16x slot to 8x but I'm ok with that, as if I used the m2.1 slot, I lose 2 SATA ports which I kind of need...

In passing through an entire controller from my motherboard to my Windows VM, I have just tried to use a USB to 2.5 inch SATA adapter for the first time, this works for a couple of minutes then windows reports the device has malfunctioned and disables it it with the reason device descriptor invalid. I've tried this in another PC and it works fine, there's nothing in the VM log to give me any ideas as to what's going on unfortunately. Has anyone else had this issue?

Looks like an SSL error, not sure how to fix that one, can you get to the GUI in safe mode?

Oh yeah, I forgot to say it's a Fractal Define R5 case.

There is no real improvement in feature support by going for anything higher than a 1050ti, the next encoder quality jump was the RTX series.

Try downloading this https://www.tightvnc.com/download/1.3.10/tightvnc-1.3.10_x86_viewer.zip Then use it to connect connect to yourunraid_ip port 5900

Something must happen when you click the VM and choose start? Error message, green arrow?

Edit the share and change Allocation Method to most free.

<source file='/mnt/user/domains/Windows 10/vdisk1.img' index='3'/> Is that the right path for the vdisk? Everything else seems to refer to Windows 101 instead.

Why do you want it to work like that out of interest?

So you don't want your disks to fill up beyond 95%?

Ryzen 5 2600 16GB DDR4 2666mhz Crucial Ballistix AT Asus TUF B450-Plus Gaming 512GB Intel NVME for Appdata/dockers Dell H310 PERC Be Quiet 750W Power Zone Power Supply Unit Silverstone SST-FS303B 2*5.25 to 3*3.5 bay adapter Additional front intake fan Fractal Design Dynamic GP-14 Fractal Define R5 Case Array: 3*3TB (WD3001FYYG) 3*6TB (WDC_WD60EZRZ) 3*2TB (TOSHIBA_DT01ACA200_63NKVAMKS) 1*2TB (WDC_WD20EZRX) 29TB usable space. Unassigned (for downloads) 1*2TB (ST32000444SS) MSI GT710 2GB DDR3 Passthrough to Windows 10 VM This picture is after 6 months of operation, the dust filters on the front of the case really do their job!

That's what you want, now the card is correctly stubbed, use another machine to access the Unraid GUI.

With regards to your syslinux config, you don't need to add multiple vfio-pci.ids options, just separate them like this vfio-pci.ids=10de:1b06,10de:10ef Try adding video=efifb:off as well. as it prevents unraid from using the EFI framebuffer which you need for OVMF based VMs If it's the primary card in the system, you'll also need to provide a vBios in the template. If it is the primary card, the terminal output when booting just freeze at the following line if correctly stubbed pci 0000:0a:00.0: vgaarb: VGA device added: decodes=io+mem,owns=io+mem,locks=none

It looks like its losing access to the vdisk and pausing because of that. You also have a misbehaving USB device, the vdisks aren't held on an external hard drive are they?

I'm curious about this too as it would free up a pcie slot on my motherboard for something else.