I’ve been setting up Truenas since the beginning of July and now everything and every service is working as intended.
My last attempt was to set up a Windows VM with HDD passthrough, so I chose a PCIe “SATA passthrough for the B550 chipset”. I shut down Truenas to connect the new HDD and tried to reboot.
Truenas now boots fine until I see the interface IP and for a few seconds the web interface.
Then I think the VM is starting and passing through every single drive, even the boot SSd for Truenas itself.
After ~2sec the web interface is gone and no access to the Truenas is possible.
I’ve attached a picture of the console. After the last line the nothing happens any more.
I think the VM has gone through every sata connection, even the boot device directly from the controller, but have not found a way to prevent this. Not with the debugg mode or anything.
Is there anything I can do instead of reinstalling Truenas itself?
SetUp:
two Storgae Pools
1 Vm Debian for PiHole
1 VM Debian for Jellyfin
1 VM HAOS
1 VM windows 11 23_H2 with Pcie Sata Controller Passthough
TrueNAS Scale Dragonfish
Hardware. R5 3500X,
32GB Ddr4 ECC Samsung,
Gigabyte Mc12 Le0,
Truenas Boot on a Samsung 860 EVO
1 NVME storagepool Samsun 960 Evo
2 3TB Seagate Baracude 3TB HHD for testing until the Toshiba Enterpise arrives.
I don’t know of an official way to prevent the VM from starting but… what if… you disabled VT-D/VT-X and/or IOMMU separation in your BIOS. Perhaps that would prevent the VMs from getting anywhere allowing you the tranquility to login and remove the passthrough?
Obviously I have not tested this, your mileage may vary.
The system is now stable again, but the GUI settings for the VMs are not loading (Error virtualisation ist not supported) and therefore cannot be changed in the web interface.
But there must be a way to change it. I’ll reseach for that
so i renamed the Zvol from „Virtual Machines“ to „Trash“ while unmounted it before.
When enabling virtualisation again i get again to the web interface Settings on the sever but after that get new error messages but still ending in I/O failure which crashes the web interface.
Maybe just reinstalling TrueNas makes the most sense.
Is there any way to safe the stable VMs with the shell? As far as i read everything is stored in the virtual Storage. Also the configuration in Truenas it self? Cause i didnt found anything to import a VM.
