timg11

@Frank1940 thanks for the background. This is the first time that I've found something that is easy in Windows, and not possible in Linux. Usually it's the other way around. ;-) I was trying to keep the sharing paths identical for existing scripts on the clients that write their images. If I have to modify those scripts I might as well just change them to \\d-root\images and avoid the complexity of the symlinks. Another thing that Windows does well is allow different permissions for nested shares. For example I could give read/write permission to \\IMAGES to all the clients, but limit access to \\d-root to a few, or make it read only. When it comes time to make an off-site backup, it is simple to mirror d-root to the external drive and it contains IMAGES and other directories.

@trurl I have a Debian VM for "normal" general purpose applications. I was thinking that something like backing up the Unraid OS config would be best done in Unraid from a cron job. Otherwise I would have to share the /boot directory on the USB drive so it could be accessed from the VM. I'll try the user scripts plugin for that task, and write the saved config to both the unraid array, as well as a separate network device.

I finally got this working. I have the passthrough set up to bind to vfio in system devices, checked the [102b:0536] 03:00.0 VGA compatible controller in IOMMU Group 20. I disabled the VGA and restarted. This succeeded without crashing Unraid. Then I enabled the VGA and it was detected as Microsoft Basic Adapter. That seemed to work, but only supported one mode 1200x800, which apparently the connected monitor does not support so there was no display. I then installed the Dell Matrox drivers for the VGA instead of the Basic Adapter drivers, and re-started the VM. Then I was able to set the resolution for the physical VGA and have an image on the monitor, along with the QXL virtual display for VNC.

Thanks! I first removed the previous symlink and re-created with the trailing slash on the first path (the target) . I may have additional complications because the share created in unraid GUI is /mnt/user/images, but the folder I'm trying to share is /mnt/user/d-root/IMAGES/ I unlinked the previous share and tried again like this: root@T440:/mnt/user/images# ls root@T440:/mnt/user/images# ln -s /mnt/user/d-root/IMAGES/ /mnt/user/images root@T440:/mnt/user/images# ls -al total 0 drwxrwxrwx 1 nobody users 28 May 31 10:51 ./ drwxrwxrwx 1 nobody users 88 May 31 10:50 ../ lrwxrwxrwx 1 root root 19 May 31 10:51 IMAGES -> /mnt/user/d-root/IMAGES// This still results in the network share "images" containing a single directory "IMAGES". I want the share to show the content of /mnt/user/d-root/IMAGES/ Edit - I also tried the trailing / on the symlink path but not the target: root@T440:/mnt/user/images# ln -s /mnt/user/d-root/IMAGES /mnt/user/images/ root@T440:/mnt/user/images# ls -al total 0 drwxrwxrwx 1 nobody users 28 May 31 11:02 ./ drwxrwxrwx 1 nobody users 88 May 31 10:50 ../ lrwxrwxrwx 1 root root 18 May 31 11:02 IMAGES -> /mnt/user/d-root/IMAGES/ root@T440:/mnt/user/images# Same result - the share contains the directory IMAGES.

I'm migrating a Windows server's sharing environment to Unraid (7.0.1 currently) Windows allows any folder to be shared as a unique network share name regardless of whether the whole drive is also shared. I want to match the structure and contents of the shares so the client systems will be able to easily migrate to the Unraid server. Here's one example of a nested sharing structure. I have copied the Windows D drive to a share on Unraid called d-root. It contains some directories that need to be shared also. d-root (shared as d) IMAGES (shared as IMAGES) IFL_temp (shared as IFL_temp) It turns out this is non-obvious to do on Unraid. I puzzeled over the Add Shares page for a while, and then started searching. I found a thread here that explains how to do it with symlinks. Thanks to @Frank1940 for the help! However I'm finding it doesn't work quite as expected. I follow this process: 1) Using the Unraid GUI for Shares, create a new share for IMAGES 2) Verify that /mnt/user/IMAGES has been created. 3) issue the command from Unraid terminal: ln -s /mnt/user/d-root/IMAGES /mnt/user/IMAGES I was expecting that the contents of /mnt/user/d-root/IMAGES would appear in /mnt/user/IMAGES What happens is that /mnt/user/IMAGES contains IMAGES@ which is the link: root@T440:/mnt/user/IMAGES# ls -l total 0 lrwxrwxrwx 1 root root 18 May 30 16:59 IMAGES -> /mnt/user/d-root/IMAGES/ From a client on the network, the share "IMAGES " contains a single directory "IMAGES" which then contains the expected contents of IMAGES. How can I make the contents of /mnt/user/d-root/IMAGES appear as the content of the network share called IMAGES ?

@trurl good point about storing the backup on the array. I understand one of the advantages of Unraid is that the disks are readable in isolation, but it would be a pain to figure out which drive contained the backup, removed it, mount it on another machine, recover it, etc. I'll find another place for it on a different system. I just found out the hard way why you say "You should use the User Scripts plugin to manage your scripts." I rebooted the Unraid system, and my /root/scripts directory and the scripts I just created were gone. root@T440:~# ls -l total 0 root@T440:~# I'll get the plugin, but Is there a guide to user scripting and how to create persistent storage for scripts in the Unraid shell? I installed "User Scripts - A plugin to act as a simple front end for any user scripts to allow you to run them without entering the command line" I'm not sure where to start, but I don't mind using the command line, I just want my scripts to remain between boots. Do I have to use the GUI, or can I still use the shell through SSH?

I turns out that anytime I stop the Windows VM, it dirty restarts the entire UnRaid system. I've disabled auto-run on the VM, and from Device Manager in the VM, disabled the VGA driver. I am now able to shut down the VM without taking down Unraid. But I still need to figure out why the VGA is not working properly. Thanks to the help of @SimonF, I'm trying the approach of bind to vfio in system devices, so I removed the hostdev XML code, and checked the [102b:0536] 03:00.0 VGA compatible controller in IOMMU Group 20. Any other suggestions to prevent full crashes of UnRaid, and enable VGA output from the Windows VM?

Thank you for showing me how the system devices are done. As I noted in my other thread, I'm also having trouble with passthrough of an embedded VGA controller. Now that I know how and where to bind to vfio in system devices, I'll remove the hostdev XML code and try the bind checkmark on the VGA controller.

I'm new to Unraid, and bringing up a 7.0.1 system. I'd like to create a simple backup script to save the /boot/config files periodically to a directory in the array. I thought a script like this could be run as a cron job daily to back up Unraid to the array: #!/bin/bash # back up flash drive unraid boot to array # NOW=$( date '+%F' ) BAKFILE=/mnt/disk1/d/IMAGES/T440-Unraid/boot-config-backup-$NOW.zip zip -r $BAKFILE /boot/config/ The problem is that my share "d" is spread across /mnt/disk1 through /mnt/disk4. Is there a preferred way to directly access a share from the Unraid shell and deterministically access a specific directory in the array? Is this the purpose of /mnt/user? it looks like it mirrors the array sharing structure.

I'm migrating a Windows server system to Unraid (7.0.1 on Dell T440) I have a 4 drive external USB3 docking station that I use for swapping drives to off site backup. These will not be part of an array, but mounted as individual drives. Portions of the main Unraid array will be synchronized to these offsite backup drives periodically. I would prefer to make these drives directly controlled by Unraid, rather than make them local to a VM. With Windows, there is a process to unmount and remove active drives. In Linux there is umount. In Unraid, for a non-array drive, would unmounting it for power down and removal be as simple as $ sudo umount /dev/sdc1 Or are there other factors to consider in the Unraid environment?

Running Unraid 7.0.1 on Dell T440 The T440 has a built-in VGA. I want to use the physical VGA for a dedicated display, but continue to use the virtual display as the main desktop for managing the system. The virtual display works fine with VNC, but I'm having trouble activating the physical VGA. As the T440 boots, the monitor attached to VGA displays the BIOS and Unraid boot process up to a shell prompt. At that time I start a browser interface. I have this in the Windows VM configuration Graphics Card: Virtual VM Console Video Driver: QXL (best) 2nd Graphics Card: Matrox Electronics 2nd Graphics ROM BIOS: (empty) I have hostdev passthrough code for the VGA controller: IOMMU: [102b:0536] 03:00.0 VGA compatible controller: Matrox Electronics Systems Ltd. Integrated Matrox G200eW3 Graphics Controller (rev 04) <hostdev mode='subsystem' type='pci' managed='yes'> <driver name='vfio'/> <source> <address domain='0x0000' bus='0x03' slot='0x0' function='0x0'/> </source> </hostdev> In the Windows VM, Device Manager shows the VGA as "Microsoft Basic Display Adapter". "This Device is working properly" I tried loading the Dell-provided drivers for the embedded Matrox video (g2w_wddm_v4_05_00_005_whql_ZPE.exe) That actually crashed the VM and Unraid. After restarting unraid system and the VM, Windows Settings / System / Display shows two monitors. Advanced display shows one display is QXL0001 and the other is Wired Display. Properties for Wired display are 1280x800, 8 bit. Everything looks normal, except there is no signal on the VGA. The monitor (that showed the Unraid boot), just goes to sleep once the Windows VM starts.

Success! The XML with <address domain='0x0000' bus='0x01' slot='0x0' function='0x0'/> as shown above works. Inserting the XML between </video> and <memballoon model='virtio'> works.

Using Unraid 7.0.1, Firefox 138.0.4 (64-bit) on Windows From the Unraid / VMs / Edit / XML view, the XML cannot be copied from Firefox to other applications. The XML can be selected and appears to be copied to the clipboard. When pasting the clipboard to another application, only a hyphen "-" character is pasted. Copy paste work as expected with Chrome. Is this a known issue? Anyone else use Firefox? Should I post a bug report?

@SimonF thanks for the reply. I don't understand "bind to vfio in system devices". I searched on that phrase with the term unraid, and found only one article that didn't explain it. On the Unraid VM Edit page, I don't find "system devices" in the form view, or in the XML. I think I understand your second suggestion, though. You're saying because the IOMMU of 00:1c.0 is the PCI bridge, I should go down one level to the USB controller which is 01:00.0 root@T440:~# readlink /sys/bus/usb/devices/usb3 ../../../devices/pci0000:00/0000:00:1c.0/0000:01:00.0/usb3 root@T440:~# readlink /sys/bus/usb/devices/usb4 ../../../devices/pci0000:00/0000:00:1c.0/0000:01:00.0/usb4 That would make my hostdev section like this: <hostdev mode='subsystem' type='pci' managed='yes'> <driver name='vfio'/> <source> <address domain='0x0000' bus='0x01' slot='0x0' function='0x0'/> </source> </hostdev> For my original question about the location of the hostdev section, I'm guessing the way I inserted it in the XML above is correct? With the changes in bus/slot/function it should work now?

BTW, this is Unraid 7.0.1 on a Dell T440 server I thought I'd try inserting the code into the XML, so I inserted it in <devices> next to last. (see below for full XML) I got this error when attempting to start the VM: Execution errorinternal error: Non-endpoint PCI devices cannot be assigned to guests Not sure if that is due to the location of the hostdev section or something else. Background information: root@T440:~# lspci | grep USB 00:14.0 USB controller: Intel Corporation C620 Series Chipset Family USB 3.0 xHCI Controller (rev 09) 01:00.0 USB controller: ASMedia Technology Inc. ASM2142/ASM3142 USB 3.1 Host Controller Which PCI is each USB bus on? root@T440:~# readlink /sys/bus/usb/devices/usb1 ../../../devices/pci0000:00/0000:00:14.0/usb1 root@T440:~# readlink /sys/bus/usb/devices/usb2 ../../../devices/pci0000:00/0000:00:14.0/usb2 root@T440:~# readlink /sys/bus/usb/devices/usb3 ../../../devices/pci0000:00/0000:00:1c.0/0000:01:00.0/usb3 root@T440:~# readlink /sys/bus/usb/devices/usb4 ../../../devices/pci0000:00/0000:00:1c.0/0000:01:00.0/usb4 The Inatech PCIe USB card appears as usb3 for USB2.0 device, and usb4 for USB3 devices. As a sanity check the Unraid boot USB drive is on usb2. IOMMU IOMMU group 15: [8086:a190] 00:1c.0 PCI bridge: Intel Corporation C620 Series Chipset Family PCI Express Root Port #1 (rev f9) So the Inatech PCIe USB controller is the only thing on 00:1c.0 . Is that not an endpoint? Here's the entire XML: <?xml version='1.0' encoding='UTF-8'?> <domain type='kvm'> <name>Windows 11</name> <uuid>530f19ee-2c76-bb83-fe14-b9b745c7552e</uuid> <description>192.168.27.2</description> <metadata> <vmtemplate xmlns="unraid" name="Windows 11" icon="windows11.png" os="windowstpm" webui="" storage="default"/> </metadata> <memory unit='KiB'>33554432</memory> <currentMemory unit='KiB'>8388608</currentMemory> <memoryBacking> <nosharepages/> </memoryBacking> <vcpu placement='static'>4</vcpu> <cputune> <vcpupin vcpu='0' cpuset='4'/> <vcpupin vcpu='1' cpuset='5'/> <vcpupin vcpu='2' cpuset='6'/> <vcpupin vcpu='3' cpuset='7'/> </cputune> <os> <type arch='x86_64' machine='pc-i440fx-9.1'>hvm</type> <loader readonly='yes' type='pflash'>/usr/share/qemu/ovmf-x64/OVMF_CODE-pure-efi-tpm.fd</loader> <nvram>/etc/libvirt/qemu/nvram/530f19ee-2c76-bb83-fe14-b9b745c7552e_VARS-pure-efi-tpm.fd</nvram> </os> <features> <acpi/> <apic/> <hyperv mode='custom'> <relaxed state='on'/> <vapic state='on'/> <spinlocks state='on' retries='8191'/> <vpindex state='on'/> <synic state='on'/> <stimer state='on'/> <vendor_id state='on' value='none'/> </hyperv> </features> <cpu mode='host-passthrough' check='none' migratable='on'> <topology sockets='1' dies='1' clusters='1' cores='2' threads='2'/> <cache mode='passthrough'/> </cpu> <clock offset='localtime'> <timer name='hpet' present='no'/> <timer name='hypervclock' present='yes'/> </clock> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>restart</on_crash> <devices> <emulator>/usr/local/sbin/qemu</emulator> <disk type='file' device='disk'> <driver name='qemu' type='raw' cache='writeback' discard='unmap'/> <source file='/mnt/user/domains/Windows 11/vdisk1.img'/> <target dev='hdc' bus='sata' rotation_rate='1'/> <serial>vdisk1</serial> <boot order='1'/> <address type='drive' controller='0' bus='0' target='0' unit='2'/> </disk> <disk type='file' device='cdrom'> <driver name='qemu' type='raw'/> <source file='/mnt/user/isos/Win11_24H2_English_x64.iso'/> <target dev='hda' bus='ide'/> <readonly/> <boot order='2'/> <address type='drive' controller='0' bus='0' target='0' unit='0'/> </disk> <disk type='file' device='cdrom'> <driver name='qemu' type='raw'/> <source file='/mnt/user/isos/virtio-win-0.1.271-1.iso'/> <target dev='hdb' bus='ide'/> <readonly/> <address type='drive' controller='0' bus='0' target='0' unit='1'/> </disk> <controller type='pci' index='0' model='pci-root'/> <controller type='ide' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x1'/> </controller> <controller type='sata' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </controller> <controller type='virtio-serial' index='0'> <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> </controller> <controller type='usb' index='0' model='qemu-xhci' ports='15'> <address type='pci' domain='0x0000' bus='0x00' slot='0x07' function='0x0'/> </controller> <interface type='bridge'> <mac address='52:54:00:c9:68:53'/> <source bridge='br0'/> <model type='virtio-net'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x02' function='0x0'/> </interface> <serial type='pty'> <target type='isa-serial' port='0'> <model name='isa-serial'/> </target> </serial> <console type='pty'> <target type='serial' port='0'/> </console> <channel type='unix'> <target type='virtio' name='org.qemu.guest_agent.0'/> <address type='virtio-serial' controller='0' bus='0' port='1'/> </channel> <channel type='qemu-vdagent'> <source> <clipboard copypaste='yes'/> <mouse mode='client'/> </source> <target type='virtio' name='com.redhat.spice.0'/> <address type='virtio-serial' controller='0' bus='0' port='2'/> </channel> <input type='tablet' bus='usb'> <address type='usb' bus='0' port='1'/> </input> <input type='mouse' bus='ps2'/> <input type='keyboard' bus='ps2'/> <tpm model='tpm-tis'> <backend type='emulator' version='2.0' persistent_state='yes'/> </tpm> <graphics type='vnc' port='-1' autoport='yes' websocket='-1' listen='0.0.0.0' sharePolicy='ignore'> <listen type='address' address='0.0.0.0'/> </graphics> <audio id='1' type='none'/> <video> <model type='qxl' ram='65536' vram='65536' vgamem='16384' heads='1' primary='yes'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x1e' function='0x0'/> </video> <hostdev mode='subsystem' type='pci' managed='yes'> <driver name='vfio'/> <source> <address domain='0x0000' bus='0x00' slot='0x1c' function='0x0'/> </source> <address type='pci' domain='0x0000' bus='0x00' slot='0x06' function='0x0'/> </hostdev> <memballoon model='virtio'> <address type='pci' domain='0x0000' bus='0x00' slot='0x05' function='0x0'/> </memballoon> </devices> </domain>

I'm following this helpful guide for USB controller passthrough. I have followed the procedure and have developed a block for code for hostdev, as explained in step 11b: <hostdev mode='subsystem' type='pci' managed='yes'> <driver name='vfio'/> <source> <address domain='0x0000' bus='0x00' slot='0x1c' function='0x0'/> </source> </hostdev> I have found the way to see the XML code by editing the VM, and using the Form View / XML View at the right. The article does not say where the new hostdev section should be inserted in the overall XML. This post has someone's complete XML with a hostdev section. In that particular file, the hostdev code goes in the <devices> section, after <video> but before <memballoon model='virtio'> Is that always the case, or dependent on other factors? Are there any details on the XML format requirements that could explain where the <hostdev> section should be inserted?

I think I figured it out. The uid was incorrect. This line in fstab appears to work: //T440.local/linux /mnt/linux cifs rw,credentials=/etc/samba/servershare.conf,uid=1000 0 0 Using UID=0 means root, which does not seem to allow R/W. UID=1000 is the first user on the Debian system, which then shows files on the share as: -rwxr-xr-x 1 debianuser root 6 May 24 11:36 test1.txt

I have Unraid 7.0.1 running on a Dell T440. I have Debian 12 running as a VM in Unraid. Debian boots off an SSD configured in Unraid as a Cache. I want the primary file system for Debian to be on the array. It all works as expected, except I do not have read/write access. I have created a Share in the array called "linux" and set it as Private under SMB I have a user for the linux VM and have given it read/write permission for the linux share. I have mounted the share in the linux vm by adding this line to fstab: //T440.local/linux /mnt/linux cifs rw,relatime,vers=3.1.1,credentials=/etc/samba/servershare.conf,uid=0 0 0 /etc/samba/servershare.conf contains the credentials. When I access /mnt/linux from the linux vm, it is read-only. I can access the linux share on the unraid array as read/write from a Windows machine. I copied some files to the linux share for testing. When I list the files in the linux share from the linux vm (ls /mnt/linux), they show user and group as "root". When I list the files from an SSH terminal in Unraid, the user is the Windows username and the group is "users". I tried sudo chown to change the files to the linux user name, but it has no effect. How can create a Linux filesystem on my Array with read/write access for my Debian 12 VM ?

Thanks for the link for backup and restore a pool. I think I could move the pool to the array, upgrade the SSDs, initialize them, then move the pool back to the SSDs. I'm not sure what Unraid would do better than the BOSS if I disable the intelligence on the BOSS S-1 card and use it as two separate drives. Is it just that Unraid does a better job of managing a RAID-1 array compared to BOSS? Unless UNRAID is doing something sophisticated, I'd think that offloading the computing for the RAID function to the BOSS card rather than the main CPU would provide some overall system-level value.

I need to add USB3 ports to my Dell T440 server running Unraid 7. In my experience, there are a variety of USB chipsets used in PCIe USB expansion cards, and some work better than others with various operating systems. Can anyone provide recommendations for PCIe-1X USB3 expansion cards known to work well with Unraid? Chipset vendors I have seen on these cards include "ASmedia", "Fresco Logic", "VLI", and the ever popular "CHINA"

So apparently the BOSS card was already configured as RAID1 in the T440 server when I installed Unraid. Here's how it was detected and now appears in Unraid: How would I give Unraid direct access? Presumably along these lines: 1) Stop VMs running on Cache device 2) create image of existing Cache Device and save onto array. (How is that done in Unraid? just cp, or is there a better tool?) 3) shut down Unraid system entirely, remove 256G SSDs and install 1T SSDs 4) boot into T440 BIOS and use BOSS configuration to make SSDs independent. 5) boot into Unraid and configure Pool Devices (presumably now there will be two) into a single Cache drive controlled by Unraid (Any guidance on that process?) 6) restore saved image on array back to cache device. 7) restart VMs

My Unraid system is running on a Dell T440 with a Dell BOSS controller containing two 256G SSDs. These are in a RAID-1 redundant array. They appear to Unraid as a single 256G volume that is a Cache Pool device. I have created two Virtual Machines with their vdisks in that volume: /mnt/user/domains/Windows 11/vdisk1.img (cache) /mnt/user/domains/Debian/vdisk1.img (cache) I'd like to replace these drives with 1TB SSDs. Is there any guidance or procedure for imaging these disks and expanding the storage? Can the controller do it itself? I.E. install one new drive, start up and the controller will replicate the remaining 256G drive onto the new 1TB drive? Then replace the other and it will replicate again? Is there any partition steps required to expand the partition to use the full size?

@Frank1940 I looked at the thread you linked, and I don't find anything about "'Share user' with a Windows Credential in Credential Manager". The string "Share user" is not in the thread. I would prefer to keep SMB signing enabled if possible. It also says "easiest way is to create a user (with a password set) in Unraid with the same name as the Windows account you are using, it should then ask you for the credentials. " That is what I am doing - I connect the Windows 11 system to Unraid using the username of the Windows 11 system. It asks for That username (tim) is also used by a different system. Them, without changing the actual username of the Windows 11 system, I used the "use a different account" option and connected to Unraid server as a different user that has an account on Unraid. Surprisingly this works. Is there some restriction about using the same user name from multiple client systems? Edit - doesn't matter - I changed the username of the Win11 system to something unique, added that user in Unraid, and now it works.

@tcream2345a, I'm having exactly the same problem on a new Unraid setup. I can reach my array shares from Windows 10, but I just added a Windows 11, and it won't accept credentials (Infinite loop of "enter network credentials"). I have my Unraid array shared with Security=Secure, Export=Yes, and case=Auto. I have users set up for each system that has access. I tried your gpedit settings on the Windows 11 system, but it hasn't fixed the issue.