SimonF

As ich777 said you need to install the gpu stats plugin + the corresponding driver package which you already have installed.

Maybe because you are using the older version of GPU Stats that is not 100% compatible with 6.12+ You need to be using my version from CA.

If you are not using PCI passthru you can either leave blank or remove. Changes where made to stop autostart if vfio binding fails as could indicate PCI numbers have chaged due to hardware removal/additions etc. If the VM autostarts and now as the wrong device is allocated to the VM due to PCI number changes it could lookup or crash the host. Before 07:00.0 VGA Card(Passthru to VM) 08:00.0 Disk Controller(Unraid) After 07:00.0 DIsk Controller(Unraid) If the VM autostarts the diskcontroller would be allocated to the VM and all the Unraid disks would not been seen on the host as a simple example.

Not sure if it is related, but you seem to be getting lots of connect errors in the logs. But you dont seem to have it installed. I cannot see any issues so far in the logs for the USB/Vm issue. It may be work disabling bluetooth on the host.

6.12.8 is on QEMU 7.2, 8+ will be in 6.13

I thought I had implemented but will look over the weekend. Normally when a usb device disconnects it is automatically removed from the guest.

Looks like you had a 2060 bound to vfio on 1:00.x which is no longer there. and a device at 09:00.0 if you don't need them delete the vfio file and next reboot you should be ok.

1024Gb is 1Tb? But it will start as a sparse file so the vdisk on the physical disk will be smaller until data is put into it.

Could you post diagnostics

Unraid uses the php api. I can look at managed save as a option which is hibernate to disk libvirt_domain_suspend libvirt_domain_suspend($res) [Since version 0.4.1(-1)] Function is used to suspend the domain identified by it's resource. @res [resource]:libvirt domain resource, e.g. from libvirt_domain_lookup_by_*() Returns:: TRUE for success, FALSE on error libvirt_domain_managedsave libvirt_domain_managedsave($res) [Since version 0.4.1(-1)] Function is used to managed save the domain (domain was unloaded from memory and it state saved to disk) identified by it's resource. @res [resource]:libvirt domain resource, e.g. from libvirt_domain_lookup_by_*() Returns:: TRUE for success, FALSE on error

Change in settings from yes to know on the VM Manager settings. if status is running still this means that Libvirt is still running, But will sort after next reboot. Or you can just change to Yes and then no and apply.

Service needs to be disabled to stop libvirt and remote the VMS opions.

I am lookingto see how this can be implemented, the rust version in the next released is stored in /usr/bin and we have a symlink from /usr/libexec to the new binary. So we could change the symlink to a wrapper to do the additional processing. I have also asked libvirt team if they have a time scale to implement in XML. Looking at the old args we should be able to build new args as follows. Old --fd=45 -o source=/mnt/cache,cache=always,sandbox=chroot,xattr /usr/libexec/virtiofsd \ --fd="$FD" \ # From fd= --shared-dir="/mnt/user/Music" \ # From source= --xattr \ # From sandbox --cache="never" \ #From cache= --sandbox="chroot" \ #From sandbox --inode-file-handles="mandatory" \ # Add as default --announce-submounts #Add as default. @jchIs there a reason for cache = "never" ?

Released as 2024.03.16

Yes I have upto 4 on my machine. They each show as a different panel. You need to be on 6.12+ and using my version of the plugin.

Would you be able to change to have this one in your PR before I merge ^

No these scripts need to be set in the sleep plug post and pre scripts. It is to disconnect the devices from the VM as they get renumbered at host resume. The disconnect just removes the attached usb devices from the VM. Is there a option to suspend as part of the sleep. I don't use sleep but these are options I provided to another user with the same issues.

You may need to use Numa 1 for the VM as that is where most of your PCI passthru devices are. I have exerimented on my system but only have 2 nodes and 64G per node. I have added this to my xml and it does remain persistent even if you update on the gui. <numatune> <memory mode='strict' nodeset='1'/> </numatune> This results in the following PID 18396 is the vm for the XML. not sure on impact if you set with 16Gb for the VM and only 16Gb is available on that node. Maybe try with 12Gb first, You can specify more that one node so you could do nodeset="1-2" or ="1,2". Per-node process memory usage (in MBs) PID Node 0 Node 1 Total --------------- ------ ------ ----- 18396 (qemu-syst 2 32478 32480 31132 (qemu-syst 1183 15529 16712 --------------- ------ ------ ----- Total 1185 48007 49192 root@unraid:~# As you have somes PCI devices on different nodes you will see IO on some of the other cores not allocated to the VM.

If you are on 6.12 you need to look at the multi versions. https://github.com/SimonFair/gpustat-unraid/blob/master/src/gpustat/usr/local/emhttp/plugins/gpustat/gpustatusmovablemulti.page

Have you found a documented best setup anywhere for the rust vers. It is included in the next release so I can look at making changes. Looking at libvirt they have not made any changes to support new options. There is an issue but at least 1 year old.

If you update via the gui it may remove those. But you can create port mappings in my plugin that will connect the devices to the VM at start without having to add to XML or template.

Yes it is possible. My plan is to add an Unraid client process that can be configurable but I have create one as yet, There are sample code called client.php which is from a pfssense plugin. here is the development guide for info: https://lcdproc.sourceforge.net/docs/current-dev.html Thanks for the icon.

Not sure if you want to test to see if this stops the error? You cannt use limit= on 6.12.8 thou. I have tested my VM runs with the new settings seems to remain persistant. But dont have the error to see if this fixes. https://github.com/tianocore/edk2/discussions/4662 <cpu mode='host-passthrough' check='none' migratable='on'> <topology sockets='1' dies='1' cores='1' threads='2'/> <cache mode='passthrough'/> <maxphysaddr mode='passthrough'/> </cpu>

I started looking at this thread, Not sure if you what to try this option? https://github.com/tianocore/edk2/discussions/4662 Currently on 6.12.8 we are running 8.7.0 Libvirt You should be able to add the maxphysaddr to the xml, and looks like it remains persistant. The limit options is currently not supported but may be on a new point release where libvirt may be bumped up. <cpu mode='host-passthrough' check='none' migratable='on'> <topology sockets='1' dies='1' cores='1' threads='2'/> <cache mode='passthrough'/> <maxphysaddr mode='passthrough'/> </cpu>