Crashes, crashes, crashes.

Case number:845829-2011835
Topic:Crash/Hang
Opened by:BootsMcGraw
Status:Open
Type:Bug
Opened on:Sunday, July 4, 2021 - 21:42
Last modified:Thursday, July 22, 2021 - 18:47

The game has crashed on me for no apparent reason three times in the last twenty-four hours. I was working different puzzles (2012, and two instances of 2014) at the time of each crash. Anyone else experiencing this? If so, chime in. If enough of you do, I'll grab a log output or whatever it is y'all say to collect when you experience a crash, and post a link to it.

(Sun, 07/04/2021 - 21:42  |  15 comments)


alcor29's picture
User offline. Last seen 11 hours 16 min ago. Offline
Joined: 11/16/2012
alcor29's picture
User offline. Last seen 11 hours 16 min ago. Offline
Joined: 11/16/2012

Also: The fact that I did not have these crashes when I had a slow computer and now that I have a fast one I crash to the point that I cannot complete any design or binder puzzle should ring a bell among all of the expert computer people we seem to have. For example, I can no longer run the script " mutate combo 1.4 for layers of these puzzles.

alcor29's picture
User offline. Last seen 11 hours 16 min ago. Offline
Joined: 11/16/2012

Hi Boots. I tried the undo settings (25/20) suggested by wheeler in https://fold.it/portal/node/2011479 It seems to be working if I get a crash still, I will try an even lower RAM setting.

alcor29's picture
User offline. Last seen 11 hours 16 min ago. Offline
Joined: 11/16/2012

Actually for tetramers, or maybe just D2s, it still crashed. Seems to better at 5% RAM

Joined: 09/24/2012
Groups: Go Science

I have the same problem with a very new computer running with Windows10 free version (the version installed to check the working of a self-made computer). i9-9900K Intel.

Window close without reason, mainly when using >1 tracks on the same Foldit map.
Very often with symmetric puzzles.
Even more if several symmetric puzzles running in parallel.
It closes during recipe run at night, or immediatly on any new action (when recipes are running on other tracks).

By chance, I saw that while running at low processor percentage, a suddend 100% processor use happened.

Joined: 09/24/2012
Groups: Go Science

Now I'm sure it's linked to this overflow of CPU on puzzel 2018. See picture, just when I opened the screen. The 3 tracks with sym puzzle had crashed (recipes: quickfix and GBF).
I had 3 couples of 2 tracks: one with puzzle 2020, one with puzzle 2018.
Only tracks 2018 crashed;
Here the log of one of the couples

Joined: 09/24/2012
Groups: Go Science

Several hours without crash with only one track: it seems that the problem might be due to working symmetric puzzles on pallel tracks (mixed with other puzzles or not). It's mainly the design puzzle track that crashes.

I have 3 copies of Foltit on my (new) computer i9. For this test, I run several tracks puzzle 2020 on one copy, several 2020 on another one, and only one track 2018 on the 3th copy. The overall CPU usage is about 75% (much higher than in the configuration above), and even if I oberve some time a 100% value, no crash occured any more running recipe Constr for several hours.

I stoped constr and ran Random Idealize : crash after a few minutes. Here is the log:

UNHANDLED EXCEPTION
[ ERROR ] Issue initializing symbols with path `C:\Foldit\cmp-binary-55def9e782733408279a65bbd90fb780;.`
[ ERROR ] Could not initialize backtrace symbols: ERR 8 Message:

On an old computer, I succeeded keeping one track 2018 in parallel with one track puzzle 2019. Recipe used is JET (a LWS recipe).

Conclusion: crash might be linked to several commands that are present in modern recipes (like idealize ? remix? or recipes that use a number of slots ?)

Joined: 09/24/2012
Groups: Go Science

Then random idealize survived some time before to crash on a single track.

Joined: 09/24/2012
Groups: Go Science

Now a crash on a track with puzzle 2020: seems due to an overload of the GPU (see image).
I saved and share a solution (uthat runned a recipe for a long time), then I deleted track, I loaded a group share, gave another name to the track and started a recipe on Med wiggle power: immediate crash.

When opening a new track, doing the same didn't repeat the crash.

Joined: 09/24/2012
Groups: Go Science

Opening again a single track. Started JET (that normally works on old computers same puzzle), same crash after some rounds:

Output recipe:
JET 4.2.6 Multiple LWS and AT with puzzle exploration, by Bruno Kestemont 103159 206056 0 0 Remaining time: 65.284166666667 Recording results in Note for segment 26 Join Evolver Team (JET 4.2.6) START Score= 26903.392 Bonus= 4200 1 night PC work. Good night Bruno Kestemont! ------------------ Stanley explorer 1.2 by BK Puzzle exploration: -There are 90 mutables; -probable free design; -There are 0 cysteins; -no possibility of bridge; -no other score; -symmetric... Pentamer Looking for a target Set of targets: 24062, 28229, 31226, 34214, 34385 Estimated top score: 31062 Current target: 31226 Distance to target = 4322 (13 %) End of Stanley exploration module ------------------ ############################ ENTREE Running Total LWS (1, 7, 0.3)... (gaining: 3.1745903350202) (gaining: 2.5522611330416)

End of Log:

interactive.application.actions.cart.PoseLoopThreadActionCart: ***** STARTING THREAD ActionLocalMinimize
interactive.application.shared.tool_util: RT: 1111111111111111111111111111111111111111111111111111111111111111111001111111111111111111111
interactive.rosetta_util.rosetta_util: Resymmetrized pose
interactive.application.actions.cart.PoseLoopThreadActionCart: ***** ENDING THREAD ActionLocalMinimize
interactive.application.shared.Tool: Time: Mon Jul 19 06:54:15 2021 UTC - Pull Mode completed
interactive.application.actions.cart.PoseLoopThreadActionCart: ***** STARTING THREAD ActionLocalMinimize
interactive.application.shared.tool_util: RT: 1111111111111111111111111111111111111111111111111111111111111111111001111111111111111111111
interactive.rosetta_util.rosetta_util: Resymmetrized pose

UNHANDLED EXCEPTION
[ ERROR ] Issue initializing symbols with path `C:\Foldit\cmp-binary-55def9e782733408279a65bbd90fb780;.`
[ ERROR ] Could not initialize backtrace symbols: ERR 8 Message:

Joined: 09/24/2012
Groups: Go Science

Remix ?

Opened again and trying quickfix. The crash occured when running remix for the second time (GPU overload recorded):

Quickfix 3.6 Option to keep freezes and bands, manage slow filters 100698 204416 0 0 Symmetric Trimer Recording results in Note for segment 26 Quickfix 3.6 Start score : 26926.754 No generic filter disabling lw = Local Wiggle id = Idealize rb = Rebuild ws = Wiggle Sidechains ws h= Wiggle worst hiding Sidechains cw= Cut Worst and wiggle all rb i= Rebuild worst ideality sh = Quick Shake rx = Remix sh h= Quick Shake worst hiding ###################### ---- LOOP 1.0 ---- wsImprovement to 26926.756 sh lw 28-30Improvement to 26926.847 lw 29-31Improvement to 26937.477 lw 30-32Improvement to 26937.518 ---- LOOP 1.1 ---- wsImprovement to 26937.523 sh lw 28-30Improvement to 26937.546 lw 29-31Improvement to 26937.624 lw 30-32Improvement to 26937.654 rx 28-31 rx 51-54

Joined: 09/24/2012
Groups: Go Science

Trying hand fold on seg 51-54 as above (24 solutions): no problem.

Running a remix recipe (Loop Remix 11), options 10 length 4. A GPU100% occured but no crash.

Stop recipe. Running again: immediate crash (GPU 100% occured).

Log with many warnings:
core.conformation.symmetry.Conformation: [ WARNING ] SymmetricConformation:: directly setting a dependent ATOM!
core.conformation.symmetry.Conformation: [ WARNING ] the jump 8 is controlled by 5
interactive.application.actions.cart.BasicLoopHashSampler: 46 fragments found for fragment size 4

UNHANDLED EXCEPTION
[ ERROR ] Issue initializing symbols with path `C:\Foldit\cmp-binary-55def9e782733408279a65bbd90fb780;.`
[ ERROR ] Could not initialize backtrace symbols: ERR 8 Message:

Startint over
Running a remix recipe (Loop Remix 11), options 10 length 4
Crash after some times. Log of the recipe here:

Loop Remix 11.0 See first comment 104760 206715 0 0 Loop remix 10.0a Score type : Normal Start score : 26944.128 Remix length : 4 Any ss Remix range 1 to 90 N remixes 10 Local mutate threshold 80 rb 1 12-15 (1/87) init score : 26944.1 (63,24) i : 1 13038.4 4840.6 157.7 81 -0.7 x -8Testing for stabilityscore with filters 26952.2score without filters 26952.2 Improvement to 26952.152 rb 2 71-74 (2/87) init score : 26952.2 (47,23) i : 1 59937.6 1400.5 1311.4 1348.3 1348.3 x x i : 2 346.3 335.7 -3.6 -3.6 -3.6 x -3.6Testing for stabilityscore with filters 26955.7score without filters 26955.7 Improvement to 26955.715 rb 3 43-46 (3/87) init score : 26955.7 (46,21)

bkoep's picture
User offline. Last seen 12 hours 46 min ago. Offline
Joined: 11/15/2012
Groups: Foldit Staff

Thanks for all of the detailed crash reports!

On July 16, we released a devprev update that fixes a crash related to the Undo Graph and autosaved solutions. If any devprev users were seeing excessive crashing, it would be helpful to know if you can repeat those crashes with the latest devprev update.

Joined: 09/24/2012
Groups: Go Science

All these where with main.
A only mutating recipe (maaa) didn't crash yet, neither GAB recipes --> it could be linked to remix or idealize functions.

Joined: 09/24/2012
Groups: Go Science

Problem solved with latest update.
No crash any more in same conditions.

Sitemap

Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons