Windows 10 guest fails shortly after startup

[techpost]

I'm having a problem with a Windows 10 Pro x64 guest on my linux (OEL 8.5) host.

The guest boots fine initially but then dies shortly after reaching the logon prompt (usually within 60s). The same host runs multiple linux guests without any problems and I see the same behaviour whether these guests are running or I shut them down first. High-level spec of the host is:
- Ryzen 5800x running on Asus Pro WS X570-ACE mobo with latest stable BIOS (4101)
- 64GB RAM
- Oracle Enterprise Linux 8.5, kernel: 5.4.17-2136.306.1.3.el8uek.x86_64
- VirtualBox 6.1.34

Guest config:

Code: Select all

Name:                        WindowsTest
Groups:                      /Test
Guest OS:                    Windows 10 (64-bit)
UUID:                        fa7f781d-a771-41fe-a02e-793090723252
Config file:                 /vm/Test/WindowsTest/WindowsTest.vbox
Snapshot folder:             /vm/Test/WindowsTest/Snapshots
Log folder:                  /vm/Test/WindowsTest/Logs
Hardware UUID:               fa7f781d-a771-41fe-a02e-793090723252
Memory size:                 8192MB
Page Fusion:                 enabled
VRAM size:                   128MB
CPU exec cap:                100%
HPET:                        enabled
CPUProfile:                  host
Chipset:                     piix3
Firmware:                    BIOS
Number of CPUs:              2
PAE:                         enabled
Long Mode:                   enabled
Triple Fault Reset:          disabled
APIC:                        enabled
X2APIC:                      enabled
Nested VT-x/AMD-V:           disabled
CPUID Portability Level:     0
CPUID overrides:             None
Boot menu mode:              message and menu
Boot Device 1:               HardDisk
Boot Device 2:               DVD
Boot Device 3:               Not Assigned
Boot Device 4:               Not Assigned
ACPI:                        enabled
IOAPIC:                      enabled
BIOS APIC mode:              APIC
Time offset:                 0ms
RTC:                         local time
Hardware Virtualization:     enabled
Nested Paging:               enabled
Large Pages:                 enabled
VT-x VPID:                   enabled
VT-x Unrestricted Exec.:     enabled
Paravirt. Provider:          Default
Effective Paravirt. Prov.:   HyperV
State:                       powered off (since 2022-05-16T20:12:50.000000000)
Graphics Controller:         VBoxVGA
Monitor count:               1
3D Acceleration:             disabled
2D Video Acceleration:       disabled
Teleporter Enabled:          disabled
Teleporter Port:             0
Teleporter Address:
Teleporter Password:
Tracing Enabled:             disabled
Allow Tracing to Access VM:  disabled
Tracing Configuration:
Autostart Enabled:           disabled
Autostart Delay:             0
Default Frontend:            headless
VM process priority:         default
Storage Controller Name (0):            SATA
Storage Controller Type (0):            IntelAhci
Storage Controller Instance Number (0): 0
Storage Controller Max Port Count (0):  30
Storage Controller Port Count (0):      4
Storage Controller Bootable (0):        on
SATA (0, 0): /vm/Test/WindowsTest/WindowsTest_Sys.vdi (UUID: 19904950-d9b5-4194-80e6-4ce209343b0d)
NIC 1:                       MAC: 08002727E34F, Attachment: Bridged Interface 'enp10s0f0', Cable connected: on, Trace: off (file: none), Type: 82540EM, Reported speed: 0 Mbps, Boot priority: 0, Promisc Policy: deny, Bandwidth group: none
NIC 2:                       disabled
NIC 3:                       disabled
NIC 4:                       disabled
NIC 5:                       disabled
NIC 6:                       disabled
NIC 7:                       disabled
NIC 8:                       disabled
Pointing Device:             USB Tablet
Keyboard Device:             PS/2 Keyboard
UART 1:                      disabled
UART 2:                      disabled
UART 3:                      disabled
UART 4:                      disabled
LPT 1:                       disabled
LPT 2:                       disabled
Audio:                       enabled (Driver: ALSA, Controller: HDA, Codec: STAC9221)
Audio playback:              enabled
Audio capture:               disabled
Clipboard Mode:              disabled
Drag and drop Mode:          disabled
VRDE:                        enabled (Address 192.168.1.2, Ports 3389, MultiConn: off, ReuseSingleConn: off, Authentication type: null)
Video redirection:           disabled
VRDE property               : TCP/Ports  = "3389"
VRDE property               : TCP/Address = "192.168.1.2"
VRDE property               : VideoChannel/Enabled = <not set>
VRDE property               : VideoChannel/Quality = <not set>
VRDE property               : VideoChannel/DownscaleProtection = <not set>
VRDE property               : Client/DisableDisplay = <not set>
VRDE property               : Client/DisableInput = <not set>
VRDE property               : Client/DisableAudio = <not set>
VRDE property               : Client/DisableUSB = <not set>
VRDE property               : Client/DisableClipboard = <not set>
VRDE property               : Client/DisableUpstreamAudio = <not set>
VRDE property               : Client/DisableRDPDR = <not set>
VRDE property               : H3DRedirect/Enabled = <not set>
VRDE property               : Security/Method = <not set>
VRDE property               : Security/ServerCertificate = <not set>
VRDE property               : Security/ServerPrivateKey = <not set>
VRDE property               : Security/CACertificate = <not set>
VRDE property               : Audio/RateCorrectionMode = <not set>
VRDE property               : Audio/LogPath = <not set>
OHCI USB:                    disabled
EHCI USB:                    disabled
xHCI USB:                    enabled

USB Device Filters:

<none>

Bandwidth groups:  <none>

Shared folders:<none>

Capturing:                   not active
Capture audio:               not active
Capture screens:             0
Capture file:                /vm/Test/WindowsTest/WindowsTest.webm
Capture dimensions:          1024x768
Capture rate:                512kbps
Capture FPS:                 25kbps
Capture options:             ac_enabled=false

Guest:

Configured memory balloon size: 0MB

If I run the same guest on my Windows 10 host, it boots perfectly and will stay up for hours.

I've attached log files from both the Linux and Windows hosts.

Things I have tried:
- Tried different number of vCPUs: 1, 2, 4 & 8
- Tried different RAM allocation: 4GB, 8GB, 16GB
- Added kernel parameter on host: "split_lock_detect=off"

Any thoughts very welcome!

[mpack]

There's no need to post logs from working guests, that only causes confusion. We know what a good log looks like, so all we need is your bad one, and a good description of the problem.

Probably unrelated, but your Guest Additions are quite old, please update.

Please also try with VRDP disabled.

At first glance this seems to a be a guest problem, not a VM problem. The VM simply detects the condition through the hypervisor API and reports it.

00:00:28.005027 GIM: HyperV: Guest indicates a fatal condition! P0=0xef P1=0xffff858fbeea72c0 P2=0x0 P3=0x0 P4=0x0
00:00:28.005052 GIMHv: BugCheck ef {ffff858fbeea72c0, 0, 0, 0}
00:00:28.005052 CRITICAL_PROCESS_DIED
00:00:29.561609 VMMDev: Guest Log: VBoxGuest: BugCheck! P0=0xef P1=0xffff858fbeea72c0 P2=0x0 P3=0x0 P4=0x0
00:00:30.248482 VMMDev: vmmDevHeartbeatFlatlinedTimer: Guest seems to be unresponsive. Last heartbeat received 4 seconds ago
00:05:45.451453 Console: Machine state changed to 'Stopping'

https://docs.microsoft.com/en-us/window ... ocess-died.

[fth0]

Some (seemingly) random ideas to pursue:

In the VM configuration, set Display > Screen > Graphics Controller to VBoxSVGA, with 3D Acceleration enabled or disabled. Note that the Windows guest OS doesn't know about the VM running headless, so the settings do make a difference.

In the VM configuration, set Audio > Host Audio Driver to PulseAudio or disable audio completely.

For a test, disable the HPET with VBoxManage modifyvm WindowsTest --hpet off.

Verify that KVM is not running simultaneously on your Linux host. VirtualBox cannot detect this reliably, leading to unexpected consequences.

Start the Windows guest OS in Safe Mode and see if the issue persists.

You could try if using a VirtualBox test build exhibits the same problem.

[techpost]

Thanks mpack and fth0 for your suggestions, have tried some experiments incorporating both of your thoughts and I think I have found a solution, here are my notes, so hopefully useful to someone else:
- I couldn't stabilise the existing guest, all of the suggestions below still resulted in the host locking up within a few seconds of boot or in a couple of cases, causing my host to kernel panic
- Decided therefore to rebuild the guest using my previous config, this was unsuccessful, guest locking up during various stages of setup
- By making the following changes to the config, I could complete setup and the guest has been stable for a number of hours
a) configured display adapter as VBoxSVGA with 128MB VRAM, 3D Acceleration disabled
b) disabled audio completely (not required by the guest)
c) left HPET enabled, the guest was less stable with it disabled
d) changed firmware to EFI

I have yet to install the current Guest Additions (unable to update on the original guest as it didn't stay up long enough) but will report back on stability after doing this.

Thanks for the pointers on reporting details and posting logs as well, helpful for next time

[fth0]

Thanks for reporting back! Can you test if the display adapter or the audio setting was the culprit?

[techpost]

Sure - no problem. I've re-enabled audio (ALSA + HD Audio controller) and the guest has been up for 1hr so will leave for the rest of the evening and try VGA next (probably Thursday I'm afraid as not around much tomorrow)

[techpost]

Hmm, so interesting test results today. Changing the VGA back to VBoxVGA has made no difference, guest was stable for 3hrs which is very good news. The less good news is that when I install Guest Additions, the guest hangs at the Windows logon screen after reboot and causes the host to kernel panic.

I'm going to perform some additional testing later today to test the same but individually rolling back the audio and vga configs - will report back on how that goes!

[techpost]

Sorry for the long delay in responding with more detail, this has taken quite a bit of testing. In summary, the problem seems to be caused by installing the guest additions. All is stable until I reboot the guest following installation at which point the host kernel panics a few seconds after the guest logon screen appears.

I tried the following config changes to try and resolve, reverting to a snapshot every time and attempting a re-install of the guest additions:
- use VBoxVGA
- use VBoxSVGA
- use VBoxSVGA with 3D acceleration enabled and 256MB VRAM (bizarrely, this didn't result in the host dying but the guest locked up during the guest addition setup)
- disable all USB controllers
- disable hpet
- disable vrde
- use ps/2 mouse rather than usb tablet (a punt in case the problem was related to the guest additions pointer management / seamless integration)

For the time-being, I'm going to leave the guest running as-is without guest additions as I generally access the guest via RDP and utilise a samba service on a linux host for shared storage so I don't lose much functionality.

If there is a forum on guest additions, maybe they have some further suggestions but I hope the above helps anybody else out with similar problems in the meantime.

[fth0]

Thanks again for reporting back!

Installing the Guest Additions rather points to the VirtualBox graphics driver, since there is no audio-related content.

[techpost]

Apologies I forgot in my last reply, I did also test with audio disabled but got the same result so that lines-up with your knowledge of the GA code.

I will raise a bug report for GA, whilst I'm not hugely concerned for this application it may help out elsewhere.

Thanks for your assistance with getting this far, I now have a usable Windows Guest!