On Tue, 15 Sep 2020 09:00:47 -0700, Mike Easter wrote:

That is a significantly mod/ed Win10.

Hi Mike,

I appreciate that you have a great grasp of the problem set.
o And that your advice is 100% on the mark

It's not a bare-bones Windows system that I'm using for the past year
o The GUI is customized to be efficient & apps are installed to be used
https://i.postimg.cc/YCZHCSnD/bsod212.jpg

I use Windows; it's not just a screensaver display; it's in use all day.
o In fact, it hasn't crashed since yesterday morning & has been stressed
https://i.postimg.cc/26yJTkFZ/voltage01.jpg

But when it BSOD'd yesterday, it took 12 pages of steps to get it back!
o https://i.postimg.cc/0yV7YFP3/bsod211.jpg

Unfortunately, I had to hone the log-saving setup after the BSOD happened:
o Win+R msconfig General Startup selection
(_)Normal startup
(o)Diagnostic startup
(_)Selective startup

Enable minidumps:
o Win+R sysdm.cpl Advanced Startup and Recovery Settings
System failure
[x]Write an event to the system log
[_]Automatically restart
Write debugging information
[none]
[Small memory dump]
[Kernel memory dump]
[Complete memory dump]
[Automatic memory dump] == select this for Win10
[Active memory dump]
Dump file:
%SystemRoot%\MEMORY.DMP
[_]Overwrite any existing file
[x]Disable automatic deletion of memory dumps when disk space is low
etc.

My point is that in a simple works vs doesn't work dichotomy, having a
simple working model as a basis of comparison to a non-working model
helps to isolate a problem.

Agreed.
o But Windows is what I use to get things done.

So "stuff" is gonna have to be added.
o In general, apps don't cause "Stop Errors" (hardware & drivers do).

In this/your scenario, we aren't even 100% confident of the hardware
integrity, except that various testing methods are negative. So, it
could be something like a bad cap that isn't visible/distorted.

There's a reason I opened this thread hoping to find a "bad cap" tester:
o What PC hardware diagnostic stress-testing freeware can you recommend?
https://groups.google.com/forum/#!topic/alt.comp.freeware/dkkdOmL95d8

So far we can only easily stress test CPUs, Memory, & GPUs (it seems).
o What we need is a "circuit board" testing suite of some sort
https://i.postimg.cc/X79V9Lv0/voltage02.jpg

Second or also, we haven't seen some 'reliable' system work for an
extended period of time, say a week. By reliable, I mean such as a live
linux or even a live H PE. Such a running clean live for a week would
give more confidence that there wasn't some kind of unpredictable
hardware condition such as a bad mobo cap.

Agreed.
o But this is my daily driver so I need to get work done.

I _tried_ to get work done on a live Ubuntu - but it was too darn slow.

But, instead, you start w/ a clean W10, then you mod it significantly
and then you get your BSOD which we don't know is caused by mobo, hdd,
or win10 + tweaks software.

Most likely, according to the cites I've already provided, a BSOD
o is usually caused by hardware about half the time, and,
o by drivers the other half of the time.

Neither one of which "should" be affected by a hundred tweaks to Windows.
o https://i.postimg.cc/D0J1tgDZ/windows-tweak.jpg

Then you have to hope that you can develop
the skills to examine Who Crashed, BlueScreenView, or Windbg and find
some answer.

Rest assured, I'm working on it, but, unfortunately, no log was saved!
o It turns out Windows is pretty damn choosy as to how it's set up.

For example...
o The page file must be on the same drive as your operating system
o The page file base allocation size must be greater than RAM
o Windows Error Reporting (WER) system service should be set to MANUAL
o Set page file to system managed on the OS drive
o Set system crash/recovery options to "kernel memory dump"
o User account control must be running.
o Sometimes SSD drives with older firmware do not create DMPS
o Cleaner applications like Ccleaner delete DMP files.
etc.

BTW, I also ran a series of de rigueur cleanups, all of which ran clean:
o For example, I ran scannow 3 times, then dism, then scannow a fourth time

From my logs (see the thread on creating a usb stick set of pc-specific
logs):
o What solution do you use to figure out what USB drive letter in a Windows
scripted command?
https://groups.google.com/forum/#!topic/alt.msdos.batch/fjxhOsMvJkY

Here's a cut-and-paste of steps from those pc-specific usb-portable logs:

o Win+R cmd {control+shift+enter}
sfc /scannow
Windows Resource Protection could not start the repair service.
sc config trustedinstaller start= auto [SC] ChangeServiceConfig SUCCESS
net start trustedinstaller
The Windows Modules Installer service is starting.
The Windows Modules Installer service was started successfully.
sfc /scannow
Beginning system scan. This process will take some time.
Beginning verification phase of system scan.
sfc /scannow
sfc /scannow
Dism /Online /Cleanup-Image /CheckHealth
DISM /Online /Cleanup-Image /RestoreHealth
sfc /scannow

I think such dump examinations are going to be exceedingly difficult.

I'm also learning how to use Microsoft SetupDiag to analyze minidump files:
o https://docs.microsoft.com/en-us/windows/deployment/upgrade/setupdiag

There is also a Microsoft Windows Performance Recorder I'm testing out:
o https://answers.microsoft.com/en-us/windows/forum/windows_10-update/windows-performance-recorder/a1648e8c-50c7-4243-9f1d-4216385c7ff3

As shown in this screenshot below:
o https://i.postimg.cc/YCZHCSnD/bsod212.jpg
--
On Usenet people help each other; in doing so, they give back to the team.