Random BSOD on new custom builds

All quick tests on both drives passed. I'll run a Long Generic on drive 0 (C:\) later today when I have three hours spare.

Ii would be a shame if the problem turned out to be AOMEI - it's by far the best free backup program I've found. I'll stop running it for a few days and disable its service.

Passed the Long Generic test.

I looked at some reviews for Macrium but didn't actually try it. If I remember rightly it's just an imaging tool: it doesn't have file restore or incremental backups.

I suggest we leave things like that for now and see what happens with AOMEI disabled. I'll report back either in a week or so or when the next BSOD occurs.

Thanks for your help so far - much appreciated.

Hi, Boozad. I hope you had a good weekend.



You'll recall that we're currently testing whether AOMEI Backupper was a problem. Time for an update...

I had a memory error reported by a program (KeePass). That needed a reboot before I could start that program again.

Then the machine locked up while it was in hybrid sleep. It woke up as scheduled but the monitor remained off and the machine was unresponsive. I had to turn it off then it resumed from the hibernation file but still didn't work properly so another reboot was required.

Then a BSOD this morning. Minidump attached just in case there's anything useful in there.

From this I conclude that AOMEI Backupper isn't the cause so I plan to re-enable it unless you object.



You'll recall that the computer supplier supplied only the hardware - all the software choices and installation are my own work.

Therefore what I really need to do is isolate whether the problem is hardware or software. I've done memtest on both machines and I've run a long test on one of the disks, all without error. I'm going to get WD's diagnostics (as all the drives are WD) and run some more tests on the disks. Are there any other hardware tests you can recommend?

What I'd normally do in this situation, given that I have two identical machines, is swap components from one to the other. But that won't help in this case as both are experiencing similar problems, and for this reason my feeling is that the cause is software.



I spent the weekend looking through the event viewer and resolving any serious issues it reported. From my research, most of them seemed to be irrelevant but at least cleaning up the dross makes it easier to see the real problems.

I'll leave it up to you to decide if you want to spend time on the second machine. They are theoretically identical and I'm reasonably confident that the problem and cause are the same on both. Looking at the second machine might give more information or just confuse the situation.



My plans - feel free to comment and/or add:
Perform any more possible tests on the hardware.
As the majority of BSODs occur at or very soon after waking from sleep for nightly maintenance I'm going to focus on that, firstly by not waking the machines for a few days.



* I recently had a problem on another machine (not one of these two). The symptoms were random hangs and crashes and everything pointed to a failing disk (not the system disk) but the diagnostics always came up clean so I looked elsewhere without success. Only when the disk failed completely was my diagnosis confirmed. So the diagnostics don't always help .

Thanks, Boozad.

I have made progress - of sorts. I've determined from the event logs that the problem isn't, as I first thought, occurring when the computer wakes from sleep. Rather, the problem is created at the time of putting the computer to sleep. It doesn't go to sleep cleanly so when it wakes up Windows thinks there's been an abnormal shutdown and produces the BSOD. So analysing the dumps isn't going to reveal anything, I think.

My next task, therefore, is to determine why it's not going to sleep cleanly, and why it's only doing it every n sleeps (where n is a number I'm hoping to determine). If you've any ideas I'd really appreciate them.

Already done. I ran Memtest for ten passes (~16 hours!) on each machine with no errors reported. I'm thinking PSU but they're a hassle to change on these machines and I first need to see if I've got a suitable replacement. I don't want to buy one because these are new machines so if the PSUs need replacing then the supplier will do it. But I first need to prove it to myself.

That said, I'm still inclining to software. I'd really like to pin down the circumstances of the failure - always the hardest part of these sort of problems :).