Windows 10 Synthetic Watchdog Timeout BSOD crash

[Nomad of Norad]

Of late, I have been getting random hangs, followed sometimes by a BSOD telling me Synthetic Watchdog Timeout. I have done just about everything I can think of to fix it. Far as I understand it, the ONLY thing that can be causing this is virtual-machines related, Hyper-V and the like. I HAD Docker Desktop installed on my machine, and some other related stuff, but I finally got totally fed up with the crashes and totally uninstalled it all, and to the best of my understanding I have turned off the Hyper-V functions from WIndows.... but tonight, I have been getting even more random hangs and crashes, just trying to watch a video in Opera, or look at chat in Discord.

This is driving me insane,

Earlier, I had completely uninstalled Docker Desktop, and then reinstalled it, and was trying to follow a detailed set of instructions on how to properly install this for the particular application I was trying to use... and that instruction had told me to install Docker Toolbox, the process of which installed some other related things from Oracle, including some sort of USB VM drivers thing. Immediately after that install was finished, several freezes and then a hang to a BSOD happened, so I immediately uninstalled all those add ons again, and then uninstalled Docker Desktop entirely.

Things went fine for a day or so, but then tonight, while watching a long video in Opera, and trying to catch up with things on a particular Discord server, the machine started hanging again, and giving me BSODs, several times. And at one point, Opera completely crashed away to desktop (basically, all its windows vanished suddenly.)

I am at my wits end here. I want all the crashes gone.

The only thing possibly related to the virtual machine left is the Pi Node desktop app I have running. It has an optional-addon side-app that runs on top of Docker Desktop, and which apparently the regular Pi Node desktop app interacts with... but with that Docker Desktop stuff removed completely, I don't really know what else could be causing this. The only other thing I can think of is, I may not have completely eradicated ALL of the pieces of that Oracle junk from my machine. o0

 

[Nomad of Norad]

Bah, it did it again, but this time all I was doing was right-click saving a bunch of JPGs from the Web to my downloads folder.

 

[Nomad of Norad]

Bah, crashed again.

 

[Nomad of Norad]

Well, I went and reinstalled back in Docker Desktop and the relevant container, and haven't had a crash since. It's pretty much what my experience was before: So long as Docker launched the container correctly, I'd only get one BSOD in a couple weeks. If for some reason Docker tripped over itself at startup of the machine, and I didn't notice it till later, then I'd more likely get these crashes. Dunno what's gong on with that.

For these past couple of days, I was trying to see if going 180 degrees the other direction, and completely uninstalling everything Docker or VM off the machine would ALSO fix the synthetic watchdog timeout issue... but it doesn't seem to have done squat in that regard, so I'm back to having Docker in place. I mainly killed out of Docker because I had run unto some sort of issue where it spontanesously one day started giving me 90% to 100% disk activity the entire day long... and one of the things I did in trying to fix that actually broke things so that the Docker container was no longer even showing up, so it couldn't launch the Pi Node app that ran on that container... and I simply got tired of trying to fix it and yanked the whole thing out.

Bah, not making THAT mistake again. 0o

Hopefully things are going to go back to not having random crashes all the time, again. And, oddly, with the reinstalled Docker and Pi Node stuff, it stopped doing that 90% to 100% disk activity all day thing. Here's hoping that stays gone, too.

 

[Nomad of Norad]

Bah, things are still locking up at random, sometimes going BAAZZZZZZZZZ for several moments (I.e. several seconds) and then resuming regular operation, sometimes going BAAZZZZZZZZZZZ for several minutes until I give up and hit Reset. Just not as often as it was a couple days ago. I can mostly go the better part of a day without it crashing, but it still shouldn't be hanging or crashing like this at ALL!

I don't understand this one jot. I haven't made any drastic changes to my machine, haven't installed or uninstalled anything else new of note, haven't added any new hardware. The only real change was various ones with Docker... and even that shouldn't have caused this magnitude of upswing in system crashes.

The apparent trigger for these crashes don't seem to make a lick of sense, either. Sometimes it goes BAZZZZ when I'm simply in the middle of watching a video on Opera browser. Sometimes it'll hang and bluescreen simply when I'm cut-pasting a file from one folder to another, or when I'm saving pictures from online to my Downloads folder, or when I'm simply typing away in chat.

It is driving me insane.

 

[kemical]

Damn, so sorry I missed your posts.. Looking now!

 

[kemical]

Hey, they all (the dumps) seem related to this watchdog timer which I'm sure we ascertained it was something to do with vmware. (as you also point out above)

Have you tried running the driver verifier? This will check if it's a driver or not?

Driver Verifier-- tracking down a mis-behaving driver.

Make sure you have a repair disk/USB in case it starts bsoding on startup, use the repair media to boot into safe mode and turn verifier off.

 

[Nomad of Norad]

Does the Driver Verifier mode take up a lot of processor-ticks, or is it low-activity enough that you barely notice?

edit: I guess, fundamentally, I'm asking: Does it hammer-hammer-hammer-hammer at the machine to make it crash so as to test it, or does it simply, quietly monitor things passively and then note things down when it crashes?

 

[Nomad of Norad]

Well, I set a restore point, then enabled the Driver Verifier thing, dutifully rebooted to let the changes take effect... and it immediately bluescreened before it could even get to desktop. I said, okay, fine, I'll go back to the restore point, and went to set a restore point from the recover-from-a-crash thingy.... and it declared there WERE no restore points. WTH?!??

The only option available -- and I made multiple attempts to get around this one -- was to let it reset-to-factory of Win10 over again, keeping my files but removing all the apps. Bah!

I've spent better part of a day getting all my usual apps and programs back in place, and moving the stuff in old Roaming to new Roaming... I've just now noticed the system briefly going BAZZZZZZ a couple times when listening to music in SL again. 0o

Would the logfiles of that stress-test still be in Windows.old somewhere?

The BSOD mentioned tracker.sys as the cause those times.

Also, at the moment I don't have any of the Docker or Pi Node stuff installed, though I did manually place the relevant folders back into Roaming.

 

[Nomad of Norad]

Okay, it kept freezing every few moments for a couple seconds at a time, and going BAZZZZ or BIZZZZZZ, but would always resume working again after it. It was doing this for over an hour. And then I walked out of the room, came back, and it had froze to BAZZZZZZZZZZ and stayed there... and it had been stuck like that for 3 minutes, judging from looking at the clock on the start-bar and at my watch. I sat there waiting for it to go on to a bluescreen, so I could do your collect-the-crash-info thing again, and it was still doing this freeze at 7 or 8 minutes.... and then suddenly the machine just completely shut off. No bluescreen, just BOOM, it was OFF!

It didn't generate any minidump, in fact it hasn't even made the minidumps folder, even though I went and turned that feature on earlier today.

Powering it back on worked normally.

What do I do from here?

 

[HenkB]

Overheated cpu?

 

[Nomad of Norad]

I don't get the impression its getting any hotter than it has been in ages.

 

[HenkB]

Clean it anyway

 

[kemical]

Well, I set a restore point, then enabled the Driver Verifier thing, dutifully rebooted to let the changes take effect... and it immediately bluescreened before it could even get to desktop. I said, okay, fine, I'll go back to the restore point, and went to set a restore point from the recover-from-a-crash thingy.... and it declared there WERE no restore points. WTH?!??

The only option available -- and I made multiple attempts to get around this one -- was to let it reset-to-factory of Win10 over again, keeping my files but removing all the apps. Bah!

I've spent better part of a day getting all my usual apps and programs back in place, and moving the stuff in old Roaming to new Roaming... I've just now noticed the system briefly going BAZZZZZZ a couple times when listening to music in SL again. 0o

Would the logfiles of that stress-test still be in Windows.old somewhere?

The BSOD mentioned tracker.sys as the cause those times.

Also, at the moment I don't have any of the Docker or Pi Node stuff installed, though I did manually place the relevant folders back into Roaming.

Apologies, I'm currently dealing with illness so apologies for not replying sooner.

I know the point is moot but this is why I said on the last post to create a rescue disk/usb.

It didn't generate any minidump, in fact it hasn't even made the minidumps folder, even though I went and turned that feature on earlier today.

From what you describe it does sound like an overheating cpu, this will sometimes appear as a bugcheck 124 machine check exception although sometimes it happens too quickly for a dump to form.

As you have reset the machine just make sure your set to produce them:
Open the run application.
Type sysdm.cpl in the run box and click ok.
Look across the top of the system properties box for 'Advanced' and click that.
Look for 'Startup and Recovery' near the bottom and click 'settings'.
Near the bottom you'll see a drop down menu under the heading 'write debugging information'.
In the drop down menu choose ' small memory dump (256KB)'
Under 'small dump directory' make sure it says %SystemRoot%\Minidump.
Click ok and your good to go.

Have you checked if your cpu thermal paste has dried out?

 

[Nomad of Norad]

I actually had made a recovery disc some while ago, but for some reason the machine would not boot from the disc, even though I have the CD/DVD drive set in BIOS as one of the things to boot from, and even though I can see that drive's light blink for a moment each time I turn the machine on, which tells me it IS in the list of devices to check for bootable media. The machine simply skipped the DVD-disc and went right to the apparently-built-in check-for-troubles-in-Windows system.

As for the sysdm.cpl thing, I did that a couple days ago.