| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Help Fight Childhood Cancer Project Forum
Thread: "Monster" Wus ? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 50
|
|
| Author |
|
|
rembertw
Senior Cruncher Belgium Joined: Nov 21, 2005 Post Count: 275 Status: Offline Project Badges: 
|
I've received 30 and 40-hour tasks too. They validate just fine. I do prefer shorter runtimes though. If something happens to a remote standing computer, less is lost due to not returning tasks on time.
Shorter tasks also facilitates the Boinc planner's job with multiple (non-WCG)projects running. |
||
|
|
kskjold
Senior Cruncher Norway Joined: May 20, 2008 Post Count: 469 Status: Offline Project Badges: 
|
My sons computer is a Dell P4 with intel M processor. And his WU has now been running for 8 hours and has only completed 16%. And Sekerob, thats not an HT processor.....:) That WU took 48 hours to complete. Is this runtime intentionally from the techs, or not? |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
No, why would that be intentional? The project average been dropping the last few days (light blue line in chart below), somewhat above the 7 hour target. Seems undulating and the lesser luck of a monster landing on your sons P4-M type computer... no way to control that at present.
---------------------------------------- Full size chart
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
nasher
Veteran Cruncher USA Joined: Dec 2, 2005 Post Count: 1423 Status: Offline Project Badges:
|
managed to get through enough of the monsters to get my emerald.. i have curently stoped crunching these till i get to my next badge issue (another blue or a new project) if the long wu's are still present i wont contunue crunchin hfcc till the next badge issue for me again.
----------------------------------------good luck geting these sorted out or curnched though  |
||
|
|
KRUMGMBH
Cruncher Germany Joined: Jul 6, 2009 Post Count: 23 Status: Offline Project Badges: 
|
I can`t say that there are monster WU`s but in the last few days my WU`s (WU`s are from the 05.12.09) takes more then 12 hours. One Week before only 7 - 8 hours.
-------------------------------------------------------------------------------- [Edit 1 times, last edit by KRUMGMBH at Dec 9, 2009 9:58:55 AM] |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Probably the below chart expresses the runtime increases better than the one linked above. Since this still early days in the s1 target of processing it might be that the techs will later cut them back to slimmer size if this holds. The daily averages have varied in past 100 days from as low as 4.67 to yesterday 8.8 hours.
---------------------------------------- Full size image link
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Somervillejudson@netscape.net
Veteran Cruncher USA Joined: May 16, 2008 Post Count: 1065 Status: Offline Project Badges: 
|
For now all HFCC
|
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Now is it me or are all the sciences experiencing longer run times... even RICE has me running 7.02 hours
---------------------------------------- (We fearlessly continue crunching)
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Hypernova
Master Cruncher Audaces Fortuna Juvat ! Vaud - Switzerland Joined: Dec 16, 2008 Post Count: 1908 Status: Offline Project Badges:
|
Here is some data of my devices. By the way they all run only HFCC at the moment.
----------------------------------------I looked at Valid WU's and the effective CPU time registered. My active devices have planet names so here we are: SATURN (Core I7 950) CPU times between 3.42 and 7.9 hrs vast majority around 5 hrs. URANUS (Core I7 950) CPU times between 8.83 and 3.25 hrs. Here 8.83 is a unique case. The vast majority is around 5 hrs. JUPITER (Core I7 975) CPU times between 2.89 and 7.51 hrs, vast majority around 4.5 hrs. VENUS (Core I7 950) CPU times between 3.12 and 8.28 hrs, vast majority around 5 hrs. SATURN and VENUS are dedicated to WCG, JUPITER and VENUS are also for personal use so work is not seamless. But JUPITER is more powerful. My initial rig was HYPERNOVA (Core2 QX9770) who has retired yesterday and was replaced by SATURN. HYPERNOVA got two WU's one at 15hrs CPU time and the other at 12hrs CPU time. What I do not understand well is why the WU that I receive on my devices were labelled 6:30 hrs. AT present they all come labelled 7:30 hrs. Is this labelling related to expected performance. Is the performance test used by WCG to calibrate the WU for a given machine? all this is not very clear for me. By the way small MERCURY is on its way. Should run hot.  |
|||
|
||||
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3716 Status: Offline Project Badges:
|
The estimated duration of a WU is a function of the amount of work to do expressed as a number of floating point operations associated to the WU and of the DCF (Duration Correction Factor) maintained in the client.
----------------------------------------For HFCC I am quite sure that the number of operations is the same for all WUs, or close to the same value anyway. The DCF is adjusted when a WU completes and is the varying factor. As you can see its value depends on what has happened for the last processed WU and not on what is expected to happen. When a WU runs longer than what was estimated by the client the DCF is adjusted "full size", i.e. queued WUs of the same project will generally be estimated at the duration of the WU which has taken longer. When a WU runs faster than estimated the DCF is decreased by only a small percentage of the difference. So it takes several normal or short WUs to bring the DCF back to what it was before a single long WU has pushed it high. If you are using a 6.10.xx version you can keep an eye on the DCF in the properties for the WCG project under the Projects tab of BOINC Manager. With the current set of applications HCMD2 is the most "disturbing" application for the DCF, and with the numbers you show HFCC must be too, although less than HCMD2. Note that the DCF is unique for a given grid, i.e. for WCG in our case and applies to all WCG applications. Cheers. Jean. PS: Do you plan to have all planets represented in your farm?  |
||
|
|
|