| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: FightAIDS@Home
Thread: Strange faah4454 Wu´s behavior |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 36
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I had two Wu´s that after they went past the 2,0% repeatedly fall back to 1,5% or so. Does somebody has an idea of what causes this strange behavior?
|
|||
|
||||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello Hollandia,
If the algorithm is non-deterministic, then this can happen. Progress can only be computed accurately for a deterministic algorithm. http://www.worldcommunitygrid.org/forums/wcg/viewthread?thread=18502#152907 Lawrence |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Thanks for the quick reply. Guess I´ll keep it running till it finds the right algorithm.
Hollandia |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
In the other explanation lawrencehardin calls it "folding", but what the program does is searching out the best docking point with the lowest possible energy required to do that. If a docking is made, on the next attempt it tries to improve and if it can't it skips forward so you'll see a typical progression of
----------------------------------------steady forward, stalling, oscillating, steady forward, stalling, oscillating etc. The graphics for BOINC are not as glossy as the one we had when the program ran on the old UD agent... the price of a uniform program that can be compiled to run on many platforms instead of only Windows. Here's the new and old side by side. The
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Oct 9, 2008 1:17:39 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Thank you for your reply.
So this `fallback` is only the programs way of finding a better or legitimate docking point? Hollandia |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
exactly!
----------------------------------------
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
HI
My name is Mauro, from Brazil. I'm a new member of World Community Grid, and I'm have the same problem with the two tasks I received from faah. I can´t complete any of them because they are always restarting, again and again. I'm not a math expert, nor in IT, to understand the process. Can you help me? Or the only way it will be to take out the faah from the projects I can help. The two tasks I'm runnning are: faah 4474 - the deadline is over faah 4498 - dead line 30/10/2008 Thanks for your help, if you can. Mauro |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
This usually means that the work unit is so large that it fails to reach a checkpoint before you restart your computer.
Should this happen again, you have two options: leave your computer on for longer, so that it can get past the difficult spot, or alternately, abort the work unit. In summary, this is caused by a number of factors coming together: a relatively slow computer, a computer that is only used for short periods, a work unit that has few checkpoints and a work unit that takes an unusually long time to run. This coincidence is rare, but it happens, as you found out. I'm sorry that you wasted your time on this work unit. |
||
|
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3715 Status: Offline Project Badges: 
|
This usually means that the work unit is so large that it fails to reach a checkpoint before you restart your computer. Mauro, if this is your case you might decide to keep only the Rice and DDDT projects in your profile. These projects have very frequent checkpoints, so first they should avoid you the problem you had, and second you would never lose more than a few minutes of work when you decide to stop your computer. In addition the jobs for the Rice project always have the same runtime, whichever the speed of the computer which runs them. They are simply doing less work on a slow computer than a fast one. Currently they run 8 hours but starting sometime this week they will be running 9 hours. This characteristic is very interesting for slower machines which usually need much more time to run jobs from other projects. For DDDT (Discovering Dengue Drugs - Together) WCG is currently waiting for work from the scientists whose lab has been severely hit by Ike hurricane but we hope to receive new work pretty soon. To change your choice of projects go to your My projects page and check/uncheck the appropriate boxes without forgetting to click the Save button at the end. Cheers. Jean. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Hi Mauro,
----------------------------------------If FAAH is your preferred project, you might want to read the Checkpoint Save FAQ explaining how to log the progress in the message window of BOINC manager. This assists to determine the time space between the moments the work is saved periodically and choose the right moment to close your computer. ciao
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
|