Message boards : Number crunching : Major problems with granted credit
Previous · 1 · 2
Author | Message |
---|---|
mikey Send message Joined: 5 Jan 06 Posts: 1895 Credit: 9,188,754 RAC: 3,501 |
If the scheduler or whatever program that stores the info about our systems could be made intelligent enough to read through the database of systems and their settings and say..oh..heres a system with 6 hour run times or longer, lets send a large model protien to it and then the same with lower run times and memory etc. Berkeley played with doing this type of thing, for units that were to be crunched and returned more quickly, but gave up on it. They said, at the time, that the way Boinc checks the computers isn't accurate enough. This was SEVERAL years ago. |
Steven Pletsch Send message Joined: 17 Oct 07 Posts: 17 Credit: 282,298 RAC: 0 |
I too have been noticing very erratic credits, this is a short list of some recent ones.. For most WU's the machines seem to be getting about 55 - 60% of their claimed value, but for others it's a bit higher than claimed. 220590191 200971180 13 Jan 2009 16:01:57 UTC 14 Jan 2009 14:20:30 UTC Over Success Done 43,126.96 177.82 254.71 220586130 200968009 13 Jan 2009 15:40:28 UTC 15 Jan 2009 3:55:44 UTC Over Success Done 42,818.62 123.67 78.32 220584480 199077349 13 Jan 2009 15:32:11 UTC 14 Jan 2009 19:22:20 UTC Over Success Done 42,572.98 202.28 218.84 220583761 200966362 13 Jan 2009 15:28:02 UTC 15 Jan 2009 3:44:36 UTC Over Success Done 43,202.18 205.27 246.02 220899883 201265495 14 Jan 2009 19:26:40 UTC 15 Jan 2009 12:27:31 UTC Over Success Done 42,370.70 111.30 69.84 220627702 201006021 13 Jan 2009 19:16:54 UTC 14 Jan 2009 23:21:06 UTC Over Success Done 41,901.18 99.97 57.82 "Every passing hour brings the Solar System forty-three thousand miles closer to Globular Cluster M13 in Hercules -- and still there are some misfits who insist that there is no such thing as progress." - Kurt Vonnegut |
Greg_BE Send message Joined: 30 May 06 Posts: 5691 Credit: 5,859,226 RAC: 0 |
I too have been noticing very erratic credits, this is a short list of some recent ones.. don't worry to much about the higher credits. that just means your machine processed those tasks faster than others crunching the same protein fragment. the opposite is true for the lower credit, your machine was not able to crunch that protein fragment as fast as other machines. in the end on your RAC it will even out. it looks like they all ran the full run time with no errors which is a good thing. ive grown used to seeing 30-40 point differences to the negative on some of the tasks. it stinks, but that's just the luck of the draw. usually due to the protein being large and complex. just remember they will all even out on your RAC. |
Message boards :
Number crunching :
Major problems with granted credit
©2024 University of Washington
https://www.bakerlab.org