Event 129 secnvme

[neatchee]

Well, I spent the evening running more tests on another fresh install of Windows and I've now got a consistent repro without ever installing PrimoCache

Sorry to bring everyone down the rabbit hole with me.

I'm going to start down the path of trying to work through the problem with ASUS (my mobo manufacturer) and Samsung

[Jaga]

Well, despite being bad news, it's also good news. Means you can begin the process of hunting for the real cause.

[neatchee]

My biggest concern is that there are so many other reported issues that can result in Event 129 - a quick Google search shows it's surprisingly common - that there might now be some other bad configuration or driver issue that I've created through my testing masking the original issue.

Only time will tell, I guess.

[neatchee]

WELL it's only been a few hours but I think I actually have this figured out finally. Haven't hit this issue since I discovered the following...

Intel Chipset Driver versions
Windows Update Provides: 10.1.1.38
Intel Driver Update Agent Provides: 10.1.1.42
Intel Website Provides: 10.1.11.3
ASUS Support Website Provides: 10.1.1.45

And just like the version from Intel's website just running the installer is not sufficient. Windows will not pick up the drivers after install.
Instead, I had to dig into the installer logs to find out where it was extracting the drivers during install and then manually update all of the Intel Chipset devices I could find in Device Manager.

I'll update this post in a few days if I still haven't been able to reproduce the issue, just to confirm.

WHAT A SAGA >.<

EDIT: Not fixed MUCH better - it took many, many hours before I saw an issue - but not actually fixed. It's manifesting slightly differently too. It looks like this time it was specifically induced by entering the Sleep state. This all leads me back to thinking it has something to do with power state handling. Gonna keep looking into different power state handling options in the Windows Registry.

[Support]

@neatchee, again, thank you very much for so many tests and sharing the results. We are glad that PrimoCache is not the cause of this issue, which is also confirmed by the topic initiator who had this kind of issue.
Our gut feeling on this problem is that it might be related to the Samsung nvme driver itself and Windows/motherboard power management. Maybe we need to ask for help from Samsung...

[neatchee]

I agree. There seems to be something very wrong with power state management on these drives. There are known issues with fixes/workarounds like completely disabling PCIe ASPM, DIPM/HIPM and MessageSignaledInterrupt (through hidden menus and registry tweaks)

If Romex has contacts at Samsung it'd be really cool if you could reach out Obviously not a high priority for you folks, but you'd be doing the internet-at-large a service

[Support]

Unfortunately we don't have contacts at Samsung tech support. We tried to call Samsung local custom service, but they focus on product quality and have no ability to solve this technical issue...
Not sure if their community can help...https://us.community.samsung.com/t5/Com ... -computing

[neatchee]

I had some additional success by manually updating the Intel ME firmware that is included in the BIOS to the last available release for the z270 chipset.
Instructions here: https://www.win-raid.com/t596f39-Intel- ... Tools.html
WARNING: Read instructions VERY CAREFULLY. Read the WHOLE POST before proceeding! Flashing the wrong Intel ME firmware can brick your system!!!!

Additionally, I exposed some additional Power Options settings including and set them to 0 (disabled):
- NVMe Idle Timeout [ Computer\HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Power\PowerSettings\0012ee47-9041-4b5d-9b77-535fba8b1442\d639518a-e56d-4345-8af2-b9f32fb26109 ]
- NVMe Power State Transition Latency Tolerance [ Computer\HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Power\PowerSettings\0012ee47-9041-4b5d-9b77-535fba8b1442\fc95af4d-40e7-4b6d-835a-56d131dbc80e ]

I also uninstalled the Samsung NVMe driver and used the Windows-provided default NVMe controller driver.

The current status is that I can use the device normally for many hours before I hit an issue, and those seem to be tied to the system going to sleep and resuming afterwards.

I have a new motherboard from a different manufacturer arriving later today so I'll report in on the results.

@RomexSupport: Thank you for tolerating my frequent status updates. I'm basically using this thread as repository for any information I come across regarding Event 129 on Samsung 970 devices. This is now one of top results on google when searching for "event 129" so I'd like to make sure anyone else who stumbles across this thread with the same issue. Nothing more frustrating than seeing "I had the same problem but I fixed it!" with no explanation as to how heh.

[Jaga]

I have a new motherboard from a different manufacturer arriving later today so I'll report in on the results.

All things considered, that's probably the best choice for your installation.

[neatchee]

All things considered, that's probably the best choice for your installation.

Only if it actually solves my problem!
My fear now is that this is an incompatibility with Intel's Z270 chipset as a whole and the Samsung 970 series NVMe drives. I've talked to folks with both the Z370 x Samsung970 and the Z270 x Samsung960 and neither report this issue

The new board is also Z270 (since that's required if I don't want to buy a new CPU as well!) so we'll see how it goes...