Hello everyone!
I recently started using Proxmox 8.2.7 (Linux 6.8.12-3-pve) to host a few VMs at home. Unfortunately, my host has crashed three times in the past two weeks without any apparent reason. I've attached the full log from shortly before the latest crash as a text file, though this section is likely the most relevant:
The hardware in use is new and has only been in operation for two weeks:
Super Micro H13SAE-MF, AMD Ryzen 9 7900X, 2x 32GB G.Skill Ripjaws S5 DDR5-5600 DIMM CL28-34-34-89, 2x Samsung 990 Pro as RAID1 for the system and a Broadcom 9500-16i HBA that is passed trough to a TrueNAS VM.
Among approximately 15 Linux VMs, I have a Windows 11 VM, which I suspect might be causing the issue. The server ran smoothly for several days until I connected to the Windows 11 VM via RDP today and opened a browser.
Here is the Windows 11 VM config:
All firmwares are up to date (Mainboard, NVMe drives, Broadcom HBA). I’m unsure how to get closer to identifying the cause. I hope someone here can offer a tip. I’d prefer not to revert to ESXi.
Best regards,
Karsten
P.S.:
I also noticed that I repeatedly lose the network connection to the Proxmox host and all VMs for a few seconds. Each time this happens, my Mikrotik switch logs a "link down" on the port. I have already swapped the port and cable.
After the Proxmox RAID1 unexpectedly degraded one day with the following log messages, I added the suggested kernel parameters (just as a side note):
I recently started using Proxmox 8.2.7 (Linux 6.8.12-3-pve) to host a few VMs at home. Unfortunately, my host has crashed three times in the past two weeks without any apparent reason. I've attached the full log from shortly before the latest crash as a text file, though this section is likely the most relevant:
Nov 12 18:41:18 pve systemd[1]: Started 110.scope.
Nov 12 18:41:19 pve kernel: tap110i0: entered promiscuous mode
Nov 12 18:41:19 pve kernel: vmbr0: port 12(tap110i0) entered blocking state
Nov 12 18:41:19 pve kernel: vmbr0: port 12(tap110i0) entered disabled state
Nov 12 18:41:19 pve kernel: tap110i0: entered allmulticast mode
Nov 12 18:41:19 pve kernel: vmbr0: port 12(tap110i0) entered blocking state
Nov 12 18:41:19 pve kernel: vmbr0: port 12(tap110i0) entered forwarding state
Nov 12 18:41:19 pve pvedaemon[1380391]: <root@pam> end task UPIDve:001528E0:01BB1DEA:6733933E:qmstart:110:root@pam: OK
Nov 12 18:41:21 pve pvedaemon[1386802]: starting vnc proxy UPID
Nov 12 18:41:21 pve pvedaemon[1377281]: <root@pam> starting task UPID
Nov 12 18:41:32 pve kernel: kvm_pr_unimpl_wrmsr: 2 callbacks suppressed
Nov 12 18:41:32 pve kernel: kvm_amd: kvm [1386740]: vcpu0, guest rIP: 0xfffff872ef13b455 Unhandled WRMSR(0xc0010115) = 0x0
Nov 12 18:41:33 pve kernel: kvm_amd: kvm [1386740]: vcpu1, guest rIP: 0xfffff872ef13b455 Unhandled WRMSR(0xc0010115) = 0x0
Nov 12 18:41:34 pve kernel: kvm_amd: kvm [1386740]: vcpu2, guest rIP: 0xfffff872ef13b455 Unhandled WRMSR(0xc0010115) = 0x0
Nov 12 18:41:34 pve kernel: kvm_amd: kvm [1386740]: vcpu3, guest rIP: 0xfffff872ef13b455 Unhandled WRMSR(0xc0010115) = 0x0
Nov 12 18:41:34 pve kernel: kvm_amd: kvm [1386740]: vcpu4, guest rIP: 0xfffff872ef13b455 Unhandled WRMSR(0xc0010115) = 0x0
Nov 12 18:41:34 pve kernel: kvm_amd: kvm [1386740]: vcpu5, guest rIP: 0xfffff872ef13b455 Unhandled WRMSR(0xc0010115) = 0x0
Nov 12 18:41:34 pve kernel: kvm_amd: kvm [1386740]: vcpu6, guest rIP: 0xfffff872ef13b455 Unhandled WRMSR(0xc0010115) = 0x0
Nov 12 18:41:34 pve kernel: kvm_amd: kvm [1386740]: vcpu7, guest rIP: 0xfffff872ef13b455 Unhandled WRMSR(0xc0010115) = 0x0
Nov 12 18:41:34 pve kernel: kvm_amd: kvm [1386740]: vcpu8, guest rIP: 0xfffff872ef13b455 Unhandled WRMSR(0xc0010115) = 0x0
Nov 12 18:41:34 pve kernel: kvm_amd: kvm [1386740]: vcpu9, guest rIP: 0xfffff872ef13b455 Unhandled WRMSR(0xc0010115) = 0x0
-- Reboot --
Nov 12 18:44:51 pve kernel: Linux version 6.8.12-3-pve (build@proxmox) (gcc (Debian 12.2.0-14) 12.2.0, GNU ld (GNU Binutils for Debian) 2.40) #1 SMP PREEMPT_DYNAMIC PMX 6.8.12-3 (2024-10-23T11:41Z) ()
Nov 12 18:44:51 pve kernel: Command line: initrd=\EFI\proxmox\6.8.12-3-pve\initrd.img-6.8.12-3-pve root=ZFS=rpool/ROOT/pve-1 boot=zfs quiet iommu=pt nvme_core.default_ps_max_latency_us=0 pcie_aspm=off
Nov 12 18:44:51 pve kernel: KERNEL supported cpus:
Nov 12 18:44:51 pve kernel: Intel GenuineIntel
Nov 12 18:44:51 pve kernel: AMD AuthenticAMD
Nov 12 18:44:51 pve kernel: Hygon HygonGenuine
Nov 12 18:44:51 pve kernel: Centaur CentaurHauls
Nov 12 18:44:51 pve kernel: zhaoxin Shanghai
Nov 12 18:44:51 pve kernel: BIOS-provided physical RAM map:
The hardware in use is new and has only been in operation for two weeks:
Super Micro H13SAE-MF, AMD Ryzen 9 7900X, 2x 32GB G.Skill Ripjaws S5 DDR5-5600 DIMM CL28-34-34-89, 2x Samsung 990 Pro as RAID1 for the system and a Broadcom 9500-16i HBA that is passed trough to a TrueNAS VM.
Among approximately 15 Linux VMs, I have a Windows 11 VM, which I suspect might be causing the issue. The server ran smoothly for several days until I connected to the Windows 11 VM via RDP today and opened a browser.
Here is the Windows 11 VM config:
All firmwares are up to date (Mainboard, NVMe drives, Broadcom HBA). I’m unsure how to get closer to identifying the cause. I hope someone here can offer a tip. I’d prefer not to revert to ESXi.
Best regards,
Karsten
P.S.:
I also noticed that I repeatedly lose the network connection to the Proxmox host and all VMs for a few seconds. Each time this happens, my Mikrotik switch logs a "link down" on the port. I have already swapped the port and cable.
After the Proxmox RAID1 unexpectedly degraded one day with the following log messages, I added the suggested kernel parameters (just as a side note):