v244 crashed two times, both while eating cooked bear steak

[wyrdsmyth]

Both crashes occurred fairly close together while outdoors exploring Mystery Lake. In each case I did not eat the entire steak because my calorie count was over 1800. About a half-hour later my computer had a BSOD and rebooted. Tonight is just not my night for playing TLD.

2015-07-17_225938.7z

2015-07-18_000408.7z

[wyrdsmyth]

So I tried to play again tonight, and experienced three crashes within less than an hour of play. Each time I was doing something different, the first I was inside the trapper's homestead and dropping wood, the second I was looking through inventory while in the trapper's homestead, and the third I was beside one of the cabins on the lake starting a fire. 7zipped crash folders are attached as before.

My initial thought was that the two crashes yesterday and the first two crashes today occurred while I was streaming over my network via Steam, running the game on my desktop but interacting via my Macbook Pro in another place in the house (CAT5 LAN connection). After the second crash tonight I went straight to my desktop and played again to see if I could rule that out - this time for a longer duration (you can see by the timestamps) but again experienced an unexpected crash.

After the first crash yesterday and the subsequent BSOD, I ran chkdsk on my hdd but it found no issues. I'm now preparing to run a memory diagnostic. As far as I know, aside from Windows updates earlier this week, and AMD Catalyst driver updates last week, no significant changes have occurred with this system to explain the sudden instability. I even rolled back my drivers after the second crash last night to the June 2015 beta (I seem to get better performance with those and less graphical glitches).

If anyone has any suggestions on how I might improve stability I'm willing to try, but the idea of forcing DX9 just seems wrong to me, somehow.

Update - I attached my output_log.txt file as well, just in case that helps. This chunk in the file seems relevant:

Fallback handler could not load library C:/Program Files (x86)/Steam/steamapps/common/TheLongDark/tld_Data/Mono/libc

Fallback handler could not load library C:/Program Files (x86)/Steam/steamapps/common/TheLongDark/tld_Data/Mono/.\libc

Fallback handler could not load library C:/Program Files (x86)/Steam/steamapps/common/TheLongDark/tld_Data/Mono/libc

Crash!!!

I'm not sure why Mono libraries would need to be called on a Windows system with .NET 4.5.2 installed, but I'm certainly not an expert on that subject.

2015-07-19_012508.7z

2015-07-19_013218.7z

2015-07-19_014828.7z

output_log.txt

[wyrdsmyth]

As it turns out, Memtest+ failed in the upper memory range sometime before it reached 40% (I stepped away and came back). So devs can probably disregard this thread, and I'll need to do some further testing to figure out which module is defective, or whether it's a motherboard, seating or heat related issue. But at this point the crashes appear to be related to hardware failure (grumble grumble only six months old grumble grumble).

[octavian]

Well, you had a BSOD, loboloco had one, and I had a CPU fan warning, all within the span of a weekend.

For what it’s worth, I think it may be a hardware issue exacerbated or caused by the fact that TLD is heavily CPU reliant, probably coupled with summer temperatures and dust. So before anything else I'd recommend opening the case and giving it a good clean if it needs one.

Even if the CPU heatsink looks clean dust bunnies usually form between the fins after a couple of months, even with filters on your intakes.

The recent “before Win 10” updates could be the problem, dust is just very easy to eliminate as a possible problem.

Other than that, random hardware failures; it can even be the combination of the quality of your country’s power supply, the quality/features of your PSU, coupled with summer months and extra stress on the power grid, all contributing to a power related thing.

But, like I said, dust/overheating is easiest to check for.

Maybe in the future the forum will get a special cubby-hole for non-game related technical issues/questions/whatever.

[wyrdsmyth]

I also have the memory sticks configured to run at their full rated speed as opposed to the motherboard's "safe" defaults, so that's probably pushing the temps on memory, board and CPU when everything starts getting interesting in the game (I do not, however, overclock - ever). The fact that three of the game crashes occurred indoors correlates with what you say, while the fourth occurred while a camp fire was getting started in the game, which affects light, temperature, etc.

My office is in my basement so the ambient temperature is fine, but power could potentially be an issue, especially considering the lights dim every time the AC compressor kicks on. I'm just not used to so many crashes in such short intervals from each other. Good point on the "before Win 10" updates as well, though I would have expected issues to start to show themselves on Wednesday or Thursday.

Dust shouldn't be an issue because the system got a thorough cleaning right around Father's day, but I'm planning to check that again as well. I used to repair personal computers for a living so I'm familiar with many of the common issues. I'm hoping re-seating RAM and setting the speed to motherboard defaults will result in a non-failing MemTest+ but I'm not optimistic. In my experience, when MemTest fails, the RAM is almost always bad.

[wyrdsmyth]

Another piece of the puzzle was that I forgot I had hooked up a 3.5" hdd from another system, which may have exceeded the available wattage provided by my 400W power supply while the system was under load, another source of overheating. I've unplugged that drive, swapped the memory sticks around (position 2 to position 1 and 1 to 2), dusted and set optimized defaults in the BIOS. Memtest86+ failed last night at sometime before 40%, we're now up to 46% with no failures yet. I'm crossing my fingers and letting it run all the way through before I proceed.

[wyrdsmyth]

I am happy to report my system appears to have stabilized. After completing MemTest86+ with no errors, I ran 3dMark (2013) in x64 mode all the way through without crashing, immediately followed by about 30 minutes of stable TLD gameplay including starting two camp fires, cooking and eating, spending time indoors and also outdoors. I won't give it a complete pass until I can complete an entire hour of play without crashing, but I'm relatively confident that the crashes have been resolved at this point.

[octavian]

Glad to hear

[Guest mcopeman]

I'm relatively confident that the crashes have been resolved at this point.

I'm glad to hear you got it working

Melody~

[wyrdsmyth]

I'm glad to hear you got it working :)Melody~

Thanks! I haven't had a crash since, and I've played two subsequent stints at over an hour each, so I'm a happy camper. I probably shouldn't have forced the full RAM speed in the BIOS - it's just odd that the system remained stable with that setting from January all the way until last Friday.