UNKNOWN_ENUM errors

If you're new to FAH and need help getting started or you have very basic questions, start here.

Moderators: Site Moderators, FAHC Science Team

Post Reply
BuddyWh
Posts: 1
Joined: Thu Jun 01, 2023 10:28 pm

UNKNOWN_ENUM errors

Post by BuddyWh »

I've been folding off and on for a while and only recently started paying attention to the logs so this could have been happening all along. What I've found recently is I can run several WU's with no issues and then suddenly one or two will throw some UNKNOWN_ENUM errors. Usually it concludes and collects the points but occasionally a CPU (in particular) WU will throw so many it discards the WU and gets another one. I've run stress/stability tests on memory, CPU and GPU with no issues. I even have a somewhat de-rated GPU profile I use for folding because it runs it so hard.

So question is...is it not uncommon to get occasional WU's that just don't....compute, to put it simply? Or should I be concerned there's something nefarious going on in my rig.

Thanks
BobWilliams757
Posts: 491
Joined: Fri Apr 03, 2020 2:22 pm
Hardware configuration: ASRock X370M PRO4
Ryzen 2400G APU
16 GB DDR4-3200
MSI GTX 1660 Super Gaming X

Re: UNKNOWN_ENUM errors

Post by BobWilliams757 »

It is possible to get errors due to the work unit, though at what rate I'm not sure. My understanding is that if any work unit continues to crash multiple times, it's stopped from further assignments and the researcher can take a look at it.

As for errors that don't cause it to fail, I've seen them happen but not often. Every now and then I see one pop up. I had one particular project that seemed to create more errors, as well as some fails. I finally figured out that with my GPU set at lower power settings they were more frequent, but I never figured out why. No others seemed to have issues with that project.

As for testing, my rig always passes tests including essentially torture testing running multiple tough programs/benchmarks/etc at once.... but it can still crash in F@H at times. The best bet is to keep track of any trends and see if you can figure anything out from there. Also keep in mind that at times one failed unit might trigger more problems without a reboot, I had that happen with a particular driver when I was running and integrated GPU.


Do you by chance run HFM so you can watch for trends on any failed projects?
Fold them if you get them!
Post Reply