Is it normal for a custom-built system to restart unexpectedly?
Is it normal for a custom-built system to restart unexpectedly?
Shutting off indicates either overheating – verify CPU temperatures; or the PSU is activating safeguards. A defective high-power GPU may cause the PSU to trigger its protections, or the PSU itself might be malfunctioning. Consider using another GPU or a different PSU.
The device appears to be operating smoothly, though it became unstable after a safe mode start. I managed to reach the troubleshoot section but had to restart, which brought me back to the BitLocker interface and then rebooted Windows to the desktop. It has remained stable for about fifteen minutes, and I was able to open the Event Viewer. Below are some recent administrative activities:
- Windows Hello for Business provisioning will not be launched.
- The device is identified as Microsoft Entra joined (or hybrid joined): Not Tested
- A user logged in using Microsoft Entra credentials: No
- Windows Hello for Business policy is active: Not Tested
- Windows Hello for Business post-logon provisioning is enabled: Not Tested
- The local system satisfies the Windows Hello for Business hardware requirements: Not Tested
- No connection detected via Remote Desktop: Yes
- A user certificate for on-premises authentication is enabled: Not Tested
- The machine follows no specific policy.
For additional information, refer to the provided link.
The default permission settings do not allow Local Activation rights for the COM Server application with the specified CLSID and APPID to the specified user account. This security setting can be adjusted using the Component Services administrative tool.
Function: CNetEnvironment::logProbeFailure
File: C:\temp\build\thehoff\Raccoon_MR50.78933656813\Raccoon_MR5\vpn\Agent\NetEnvironment.cpp
Line: 1751
Invoked Function: CHttpProbeAsync::SendProbe
Return Code: -27000818 (0xFE64000E)
Description: HTTP_PROBE_ASYNC_ERROR_CANNOT_CONNECT
HTTP (host: 72.163.1.80)
Function: PreferenceMgr::invokePreferenceUpdateCBs
File: C:\temp\build\thehoff\Raccoon_MR50.78933656813\Raccoon_MR5\vpn\Api\PreferenceMgr.cpp
Line: 1756
Callback interface address set is empty.
License Activation (slui.exe) encountered error:
hr=0x80004005
Command-line arguments:
RuleId=31e71c49-8da7-4a2f-ad92-45d98a1c79ba;Action=AutoActivate;AppId=55c92734-d682-4d71-983e-d6ec3f16059f;SkuId=221a02da-e2a1-4b75-864c-0a4410a33fdf;NotificationInterval=1440;Trigger=UserLogon;SessionId=1
Function: CThread::invokeRun
File: C:\temp\build\thehoff\Raccoon_MR50.78933656813\Raccoon_MR5\vpn\Common\Utility\Thread.cpp
Line: 500
Invoked Function: IRunnable::Run
Return Code: -32112629 (0xFE16000B)
Description: BROWSERPROXY_ERROR_NO_PROXY_FILE
Function: CNetEnvironment::testNetwork
File: C:\temp\build\thehoff\Raccoon_MR50.78933656813\Raccoon_MR5\vpn\Agent\NetEnvironment.cpp
Line: 1029
Invoked Function: CNetEnvironment::IsSGAccessible
Return Code: -28901363 (0xFE47000D)
Description: NETENVIRONMENT_ERROR_PROBE_INCOMPLETE:Network Probe could not reach the target
Function: CNetEnvironment::TestAccessToSG
File: C:\temp\build\thehoff\Raccoon_MR50.78933656813\Raccoon_MR5\vpn\Agent\NetEnvironment.cpp
Line: 1678
Invoked Function: CNetEnvironment::analyzeHttpResponse
Return Code: -28901363 (0xFE47000D)
Description: NETENVIRONMENT_ERROR_PROBE_INCOMPLETE:Network Probe failed to connect
Here is what I found in the reliability monitor.
The issue indicates a hardware-related fault that led Windows to malfunction.
Problem signature details include event names, code values, and parameter information.
Relevant files listed are WATCHDOG-20250531-1119.dmp, sysdata.xml, memory.csv, sysinfo.txt, and others.
Caution is advised as opening these files might damage your system if a virus or security threat was involved.
I went to MSI Afterburner just to test things out and couldn't locate the 4090 GPU, only seeing the AMD integrated card listed. I then closed Device Manager, turned off the AMD card, and everything functioned properly. My guess is that the AMD card was taking precedence over the Nvidia as the default, and the CPU was overheating. Sounds crazy?
If you don't have a sufficient cooler, it's best to turn off the iGPU via the BIOS completely.