| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: [HALP] What variable determines amount of work downloaded |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 23
|
|
| Author |
|
|
Alther
Former World Community Grid Tech United States of America Joined: Sep 30, 2004 Post Count: 414 Status: Offline Project Badges: 
|
i anticipated that question.....1 hour minimised to tray v 1 hour open / on top. Repeated several time. In all cases did not touch the machine.......i'll for sure keep it open. Probably would have todo it over longer periods as some birdy told me that a WU can have sections that fold less easy than others, but if 1 hour minimised to tray progresses the WU by 11% and expanded 13%, repeatedly, i'll discount the heavy section theorem.  Leaving the GUI open will not increase the speed of the WU. In fact, if anything, it will slow it down (albeit very, very slightly) since it has to take CPU power away to update the text and messages. You can't base your theory on percent complete since WUs vary in the length of time they take to complete. FAAH WUs are a little more predictable than HPF WUs, but they still vary. The only thing you can really compare is CPU time used. Even that has a margin of error in it since there are so many processes running on modern systems these days, you never know when a background process needs a little CPU. So, unless the amount of CPU time varies by more than this fuzzy 'margin of error', you can't claim it's faster one way or another. Some people who know Windows in depth may claim that the process which has the GUI in the foreground gains a little in terms of priority. This is true, but you must remember that the actual science app has no GUI. They are separate processes and while the BOINC Mgr will gain a little in priority, it has no effect on the science app.
Rick Alther
----------------------------------------Former World Community Grid Developer [Edit 1 times, last edit by Alther at May 15, 2006 2:32:20 PM] |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
I was basing the measurements strictly within single FAAHs. 1 hours of wallclocktime segments against % progress in BOINC minimised to tray, versus expanded on screen.
----------------------------------------Actually, i'm starting to doubt the accuracy of progress reported by BOINC GUI v CPU reported time. Sometimes the percent drags in ratio to CPU time, then is races along. I voiced this theory of poor connectivity between front and back-end. Fact, the front end crashes, yet leaving the backend process run unharmed, restarting frontend, pickup up where it left would suggest that....does the crunch part keep track of its progress or does the front end make assumption based on data points in the WU or some other algorythm based on the size of the WU? As reported, i killed all possible scheduled process i could to reduce the impact. This also to determine if any of these were infringing i.e. are connected to the GUI freezing. Suspected the Logitech/DiNovo keyboard/mouse/Bluetooth at one time. I let it rest, as the reported workarounds allow continued crunching.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Alther
Former World Community Grid Tech United States of America Joined: Sep 30, 2004 Post Count: 414 Status: Offline Project Badges:
|
I was basing the measurements strictly within single FAAHs. 1 hours of wallclocktime segments against % progress in BOINC minimised to tray, versus expanded on screen. Again, wallclock time and % progress is an inaccurate measurement. CPU time is the only accurate measurement. Actually, i'm starting to doubt the accuracy of progress reported by BOINC GUI v CPU reported time. Sometimes the percent drags in ratio to CPU time, then is races along. I voiced this theory of poor connectivity between front and back-end. Fact, the front end crashes, yet leaving the backend process run unharmed, restarting frontend, pickup up where it left would suggest that....does the crunch part keep track of its progress or does the front end make assumption based on data points in the WU or some other algorythm based on the size of the WU? The science app and the BOINC Manager are two separate processes. The science app keeps track of the percent complete (or in FAAHs case estimated percent complete) and reports this back to BOINC.
Rick Alther
Former World Community Grid Developer |
||
|
|
|