TrueNAS VM won't start after upgrade to 8.4 [Need urgent help!]

R

RealPjotr

New Member
May 31, 2023
26
12
3
I have a TrueNAS Scale VM with some disks, this is the first VM that should boot on my node. Other VMs can be started, but this one can't.

My Proxmox "iso" network storage resides on this VM. But I've run it like this for years without issues.

"qm start 108" just hangs now and the system log messages are repeated (I disabled "iso" and those messages disappear, it's not related to the VM start issue I think). I've tried a full reboot. I've tried "systemctl restart pvestatd". What can I do?

reboot hangs on this (I power cycled eventually):
1744903999925.png

system log:
Code: 
Apr 17 16:54:33 kosmos pvestatd[4490]: storage 'iso' is not online
Apr 17 16:54:33 kosmos pvestatd[4490]: status update time (6.605 seconds)

When I try "qm start 108", it also creates a "-1" lock file:
Code: 
# ls -al /var/lock/qemu-server/lock*
-rw-r--r-- 1 root root 0 Apr 17 17:08 /var/lock/qemu-server/lock-108.conf
-rw-r--r-- 1 root root 0 Apr 17 17:08 /var/lock/qemu-server/lock--1.conf

qm config 108:
Code: 
agent: 1
balloon: 0
boot: order=scsi0;ide2;net0
cores: 4
cpu: x86-64-v3
hostpci0: 0000:c8:00,pcie=1
hostpci1: 0000:c1:00.0,pcie=1
hostpci2: 0000:c2:00.0,pcie=1
ide2: none,media=cdrom
machine: q35
memory: 16384
meta: creation-qemu=9.0.2,ctime=1731770400
name: truenas3
net0: virtio=BC:24:11:98:C1:64,bridge=vmbr0,firewall=1
numa: 0
onboot: 1
ostype: l26
scsi0: spooldata:vm-108-disk-0,discard=on,iothread=1,size=32G
scsihw: virtio-scsi-single
smbios1: uuid=22720d88-7a8d-46e3-acb2-ad229c0929cb
sockets: 1
startup: order=1,up=60
tags: infra
vmgenid: 2090fdd8-2792-46e1-9d24-ea2e9894bcb7
 
Last edited:
The issue is with one of the PCIe devices: hostpci0: 0000:c8:00,pcie=1
When I remove it, the VM can boot. This is the SATA controller. Nothing has changed about it, it's still in the hardware list as before.

Any idea how to debug this issue? There is nothing in the logs...
 
Last edited:
The cause of the problem is upgrading the kernel to 6.14. I reverted to 6.11, then the VM boots again.

I'll stick to 6.11 for now. Hopefully this can help someone else.
 
I just looked at the qm config for my TrueNAS VM which came up fine after I updated and I see some things that are different. Maybe these might help you, not sure. First off, I am using a UEFI bios (OVMF) and not the legacy/default SeaBIOS. Second, my machine type is i440fx not q35. Third, my CPU type is set to host. Finally I am only passing through one PCI device, a M.2 to SATA controller. It looks like you are passing through three devices? What are the other two? Also, my SATA controller pass through reads "hostpci0: 0000:06:00" The attribute "pcie=1" is not appended on mine. I am not sure what that setting is or does.
 
You must log in or register to reply here.
Top Bottom