| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Beta Testing
Forum: Beta Test Support Forum
Thread: New Beta, C4CW v 6.13 August 24, 2010 |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 73
|
|
| Author |
|
|
Somervillejudson@netscape.net
Veteran Cruncher USA Joined: May 16, 2008 Post Count: 1065 Status: Offline Project Badges: 
|
still a spattering of Betas on my machines.
|
||
|
|
anhhai
Veteran Cruncher Joined: Mar 22, 2005 Post Count: 839 Status: Offline Project Badges: 
|
Sekerob, one of my machine recently crashed (power outage --stupid storms). And when I restarted it I noticed that one of the WU restarted at 0%, but already had over 1 hr of work. So I still don't understand how these checkpoints work (well unless by chance the system died when it was doing a checkpoint). Once it was finished I was credited for the extra hr. However, due to the fact this one had a wingman check it, I didn't get extra pts, just the extra hr or work.
----------------------------------------Here is a part of the file that you may be interested in: [03:28:19] Percent complete = 51.658057 [03:31:01] Percent complete = 53.324446 [03:33:45] Percent complete = 54.990835 [03:36:26] Percent complete = 56.657224 Commandline = projects/www.worldcommunitygrid.org/wcg_c4cw_lmps_6.13_windows_intelx86 -screen none -in in.wcg.acc -var wcgsteps1 1000 -var wcgsteps2 5000 -var loop 0 -var restart 0 -var rinterval 100 -var ifile in.wcg.acc -var wcgseed 416634146 [11:11:27] Percent complete = 1.666389 [11:13:53] Percent complete = 3.332778 [11:16:26] Percent complete = 4.999167 [11:19:02] Percent complete = 6.665556 [11:21:35] Percent complete = 8.331945 [11:24:11] Percent complete = 9.998334  |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Put it under 'client bug'. The Elapsed time you see in the BM is not produced by the science app, rather it's simply the wallclock time during which a science app is run. If a sciences resets either to the last checkpoint or HF, due unfortunately timing back to zero it's supposed to loose the associated run time. Does not seem to do so and per your observation has even led to extra hours recorded on the website. Whilst that is statistically great, it not good for client task scheduling, since when a regular task finishes in double time, all other buffered tasks are assumed to be running longer too.
----------------------------------------Regrettable, and fortunately as well, can't look over your shoulder what is happening on your rig, so let's give this a pass (and get a cheap UPS). We can't analyze every single 'Strange' event if not readily replicable. Most important the task validates. BTW, sometimes you see that extra on first look and think things are lossless, but then when BOINC has been running for a few minutes and finding back it's composure it's not new to see those extra times and %s still disappear.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
|