Hello dear people !
My system is composed from:
MB: ASROCK - EPYCD8
CPU Model: AMD EPYC 7402 24-Core (48 threads) Processor
Memory: 252 GiB
HDD Space: ~ 320 TB (main system + DAS)
3 x NVIDIA Corporation GP104GL [Tesla P4]
I have 2 bugs (until now) after the update of my TrueNas from 22.12.4.2 to 23.10.1:
Firs bug:
After the update the system has only 1 thread that is staying at 100%, after every reboot there is a different thread at 100%. I have checked all the PCI's and everything it is OK, all PCI's are working within parameters.
Seccond bug:
The 3 x NVIDIA Corporation GP104GL [Tesla P4] are isolated but i have a problem with one.
I have 3 Virtual Machines :
First is with PLEX MediaServer - first Tesla P4 allocated for hardware transcoding. (PCI ID 41:00.0)
Seccond is with Windows for Cloud Gaming - Seccond Tesla P4 (PCI ID 81:00.0)
Third is with Windows for Cloud Gaming - third Tesla P4 (PCI ID 82:00.0)
The problem is on the third one. Firs i have tryied to edit the VM to uncheck "Ensure Display Device" and i couldn't because i recieved the error message "get_pci_ids_for_gpu_isolation" not found in "device". I removed the the PCI card from Devices and tried againg. I recieve the message that the "VM updated successfully." but still the check mark stays activated on "Ensure Display Device". I have tried after that to add again the PCI card and now it shows this message "0000:82:00.0: Not Available by 'Not Available'". What is strange, is that the PCI card is isolated and if i go in main shell and list all the PCI Cards (lspci), my Tesla P4 it shows normally.
For the moment I didn't reboot the server. I will get back after the reboot.
In attachment you have the debug file.
Thank you and Happy Holidays !
My system is composed from:
MB: ASROCK - EPYCD8
CPU Model: AMD EPYC 7402 24-Core (48 threads) Processor
Memory: 252 GiB
HDD Space: ~ 320 TB (main system + DAS)
3 x NVIDIA Corporation GP104GL [Tesla P4]
I have 2 bugs (until now) after the update of my TrueNas from 22.12.4.2 to 23.10.1:
Firs bug:
After the update the system has only 1 thread that is staying at 100%, after every reboot there is a different thread at 100%. I have checked all the PCI's and everything it is OK, all PCI's are working within parameters.
Seccond bug:
The 3 x NVIDIA Corporation GP104GL [Tesla P4] are isolated but i have a problem with one.
I have 3 Virtual Machines :
First is with PLEX MediaServer - first Tesla P4 allocated for hardware transcoding. (PCI ID 41:00.0)
Seccond is with Windows for Cloud Gaming - Seccond Tesla P4 (PCI ID 81:00.0)
Third is with Windows for Cloud Gaming - third Tesla P4 (PCI ID 82:00.0)
The problem is on the third one. Firs i have tryied to edit the VM to uncheck "Ensure Display Device" and i couldn't because i recieved the error message "get_pci_ids_for_gpu_isolation" not found in "device". I removed the the PCI card from Devices and tried againg. I recieve the message that the "VM updated successfully." but still the check mark stays activated on "Ensure Display Device". I have tried after that to add again the PCI card and now it shows this message "0000:82:00.0: Not Available by 'Not Available'". What is strange, is that the PCI card is isolated and if i go in main shell and list all the PCI Cards (lspci), my Tesla P4 it shows normally.
For the moment I didn't reboot the server. I will get back after the reboot.
In attachment you have the debug file.
Thank you and Happy Holidays !