Task hang on return from Windows hibernation.

Jord
Joined: 26 Jan 05
Posts: 2,952
Credit: 5,683,412
RAC: 432
Topic 196425

A weird thing I just noticed. I only run BRP4s on my ATI HD6850 and the one that ran moments ago, was running for over 2 hours with progress stuck at 97.921%.

My system's been hibernating all day long, it switches off automatically in the morning 10 minutes after BOINC's been suspended for time-of-day. I restarted the computer at 21:24 local time and just now returned to my computer. Saw in BM that long run-time and no progress. GPU-Z showed my GPU temp was well below its normal 65C.

So I exited BOINC, waited for the OpenCL app to leave memory, then restarted BOINC.

I don't know what caused the hang, seems to be a fluke in the science app. stderr.txt says about that episode:

Activated exception handling...
[21:24:08][4896][INFO ] Output file: '../../projects/einstein.phys.uwm.edu/b2030.20110927.G48.21+01.04.S.b5s0g0.00000_2320_1_0' already exists - skipping pass
Activated exception handling...
[23:09:42][1800][INFO ] Output file: '../../projects/einstein.phys.uwm.edu/b2030.20110927.G48.21+01.04.S.b5s0g0.00000_2320_1_0' already exists - skipping pass
[23:09:42][1800][INFO ] Output file: '../../projects/einstein.phys.uwm.edu/b2030.20110927.G48.21+01.04.S.b5s0g0.00000_2320_1_1' already exists - skipping pass
[23:09:42][1800][INFO ] Output file: '../../projects/einstein.phys.uwm.edu/b2030.20110927.G48.21+01.04.S.b5s0g0.00000_2320_1_2' already exists - skipping pass
[23:09:42][1800][INFO ] Output file: '../../projects/einstein.phys.uwm.edu/b2030.20110927.G48.21+01.04.S.b5s0g0.00000_2320_1_3' already exists - skipping pass
[23:09:42][1800][INFO ] Output file: '../../projects/einstein.phys.uwm.edu/b2030.20110927.G48.21+01.04.S.b5s0g0.00000_2320_1_4' already exists - skipping pass
[23:09:42][1800][INFO ] Output file: '../../projects/einstein.phys.uwm.edu/b2030.20110927.G48.21+01.04.S.b5s0g0.00000_2320_1_5' already exists - skipping pass
[23:09:42][1800][INFO ] Output file: '../../projects/einstein.phys.uwm.edu/b2030.20110927.G48.21+01.04.S.b5s0g0.00000_2320_1_6' already exists - skipping pass
[23:09:42][1800][INFO ] Starting data processing...
[23:09:43][1800][INFO ] Using OpenCL platform provided by: Advanced Micro Devices, Inc.
[23:09:43][1800][INFO ] Using OpenCL device "Barts" by: Advanced Micro Devices, Inc.
[23:09:43][1800][INFO ] Continuing work on ../../projects/einstein.phys.uwm.edu/b2030.20110927.G48.21+01.04.S.b5s0g0.00000_2327.bin4 at template no. 4902

In the mean time the task has finished correctly. After restart it started from a previous checkpoint at 1h 10m, a lot less than the 2h+