VM Keeps Shutting down; KVM: entry failed, hardware error 0x80000021

[brelade]

Hi All,

I am new to the whole Proxmox virtualisation, so please bare with me!

We currently have a Dell r710 with x2 Xeon E5-2666 and 192gb RAM.

We are experiencing a problem where our Windows Server 2019 VM keeps shutting down at least once or twice a day.

Here's the log:
Jul 07 12:07:31 ve QEMU[4191355]: KVM: entry failed, hardware error 0x80000021 Jul 07 12:07:31 ve QEMU[4191355]: If you're running a guest on an Intel machine without unrestricted mode Jul 07 12:07:31 ve QEMU[4191355]: support, the failure can be most likely due to the guest entering an invalid Jul 07 12:07:31 ve QEMU[4191355]: state for Intel VT. For example, the guest maybe running in big real mode Jul 07 12:07:31 ve QEMU[4191355]: which is not supported on less recent Intel processors. Jul 07 12:07:31 ve kernel: set kvm_intel.dump_invalid_vmcs=1 to dump internal KVM state. Jul 07 12:07:31 ve QEMU[4191355]: EAX=00000080 EBX=00000056 ECX=00000000 EDX=a355e080 Jul 07 12:07:31 ve QEMU[4191355]: ESI=4aaedf80 EDI=b09e6c58 EBP=3c44ccc0 ESP=4c472fb0 Jul 07 12:07:31 ve QEMU[4191355]: EIP=00008000 EFL=00000002 [-------] CPL=0 II=0 A20=1 SMM=1 HLT=0 Jul 07 12:07:31 ve QEMU[4191355]: ES =0000 00000000 ffffffff 00809300 Jul 07 12:07:31 ve QEMU[4191355]: CS =7600 7ff76000 ffffffff 00809300 Jul 07 12:07:31 ve QEMU[4191355]: SS =0000 00000000 ffffffff 00809300 Jul 07 12:07:31 ve QEMU[4191355]: DS =0000 00000000 ffffffff 00809300 Jul 07 12:07:31 ve QEMU[4191355]: FS =0000 00000000 ffffffff 00809300 Jul 07 12:07:31 ve QEMU[4191355]: GS =0000 00000000 ffffffff 00809300 Jul 07 12:07:31 ve QEMU[4191355]: LDT=0000 00000000 000fffff 00000000 Jul 07 12:07:31 ve QEMU[4191355]: TR =0040 4c459000 00000067 00008b00 Jul 07 12:07:31 ve QEMU[4191355]: GDT= 4c45afb0 00000057 Jul 07 12:07:31 ve QEMU[4191355]: IDT= 00000000 00000000 Jul 07 12:07:31 ve QEMU[4191355]: CR0=00050032 CR2=04394068 CR3=10300000 CR4=00000000 Jul 07 12:07:31 ve QEMU[4191355]: DR0=0000000000000000 DR1=0000000000000000 DR2=0000000000000000 DR3=0000000000000000 Jul 07 12:07:31 ve QEMU[4191355]: DR6=00000000ffff0ff0 DR7=0000000000000400 Jul 07 12:07:31 ve QEMU[4191355]: EFER=0000000000000000 Jul 07 12:07:31 ve QEMU[4191355]: Code=kvm: ../hw/core/cpu-sysemu.c:77: cpu_asidx_from_attrs: Assertion `ret < cpu->num_ases && ret >= 0' failed. Jul 07 12:07:36 ve kernel: vmbr0: port 2(tap100i0) entered disabled state Jul 07 12:07:36 ve kernel: vmbr0: port 2(tap100i0) entered disabled state Jul 07 12:07:36 ve systemd[1]: 100.scope: Succeeded. Jul 07 12:07:36 ve kernel: zd32: p1 p2 p3 p4 Jul 07 12:07:36 ve systemd[1]: 100.scope: Consumed 4h 5min 10.536s CPU time. Jul 07 12:07:40 ve qmeventd[112263]: Starting cleanup for 100 Jul 07 12:07:40 ve qmeventd[112263]: Finished cleanup for 100

Running PVE Version
PVE 5.15.30-3

Any help would be greatly appreciated.

 

[datschlatscher]

Yes, this is a problem we have been aware of for some time now, which affects the new 5.15 kernel versions.

There is a workaround that seems to fix this issue detailed in this forum post.
For some more information on this issue in a more tidy format please also take a look at the upgrade guide PVE 6.0 -> 7.0 - KVM: entry failed, hardware error 0x80000021

 

[brelade]

Yes, this is a problem we have been aware of for some time now, which affects the new 5.15 kernel versions.

There is a workaround that seems to fix this issue detailed in this forum post.
For some more information on this issue in a more tidy format please also take a look at the upgrade guide PVE 6.0 -> 7.0 - KVM: entry failed, hardware error 0x80000021

Thank you datshlatscher! I followed the second link you sent, and so far seems to be good. I have yet to experience a KVM Error. It has been 19 hours since implementing the 'work around' so only time will tell.....

Cheers!