| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Beta Testing
Forum: Beta Test Support Forum
Thread: DDD2 Type B work units going out. |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 369
|
|
| Author |
|
|
Hypernova
Master Cruncher Audaces Fortuna Juvat ! Vaud - Switzerland Joined: Dec 16, 2008 Post Count: 1908 Status: Offline Project Badges: 
|
Just returned a valid WU this time :
----------------------------------------BETA_ erlc_ a043_ pe0000_ 0-- 613 Valid 15.01.10 23:01:08 17.01.10 07:58:01 5.95 153.0 / 152.1 My wingmen returned a Valid one and an Invalid one so quorum was reached. But I saw that the Invalid one was also granted full credit. What is interesting is that this WU when it arrived to me it was labeled 15 hrs CPU time. Finally it took me 5.95 hrs CPU time. How is the estimated CPU time calculated?  |
|||
|
||||
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
For points credit on Invalid results refer to an earlier comment I made in this thread, same as valid, but in production you get half of the quorum mean, and full add-in of the CPU time.
----------------------------------------The projects in production have a dynamic FPOP estimate in the header based on the actual averages of prior days. The BETA's have estimates in the header with a large safety, so anyone getting BETA work does not overcommit should the non-deterministic calculations, if any, work out for there to be much more. Think you might remember those BETA tests where jobs ran 100 hours+
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Am pleased to report that all 4 Beta jobs, good for 56 hours of CPU time, 2 running uninterrupted, 2 run with forced interruption validated.
----------------------------------------BETA_ erlc_ a053_ pe0000_ 1-- 95711 Valid 1/15/10 23:02:52 1/17/10 12:11:55 14.81 158.3 / 129.7 BETA_ erlc_ a051_ pe0000_ 1-- 95711 Valid 1/15/10 23:02:31 1/17/10 07:32:16 13.93 148.8 / 86.4 ** BETA_ erlc_ a049_ pe0000_ 1-- 95711 Valid 1/15/10 23:02:09 1/16/10 14:41:04 14.03 150.3 / 150.6 BETA_ erlc_ a048_ pe0000_ 0-- 95711 Valid 1/15/10 23:01:49 1/16/10 14:40:42 13.89 148.9 / 134.7 ** looking at the 2 wingmen that produced quorum before mine was returned, these were both quite substantial underclaimers, or a bit more compile tuning is needed to make they run faster on Centrino Duo's, on which claims and credit are usually very close or better than asked for. And, if anyone understand this claim spread, I'd be delighted to learn: BETA_ erlc_ a053_ pe0000_ 0-- 613 Valid 1/15/10 23:03:02 1/16/10 20:32:07 19.19 337.7 / 129.7 BETA_ erlc_ a053_ pe0000_ 1-- 613 Valid 1/15/10 23:02:52 1/17/10 12:11:55 14.81 158.3 / 129.7 BETA_ erlc_ a053_ pe0000_ 2-- 613 Valid 1/15/10 23:02:51 1/16/10 23:08:37 5.32 120.0 / 129.7 *** *** Suspected underclaimer.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
small additional amounts of type B.pe work units are happening right now. There are only 210 of them. But it is to test the new validation logic that is in place. If all goes well, then we will be releasing the type B.se work units Monday. Again, that is if all goes well. Good luck and thank you! -Uplinger Might it have been 210x3 that went out? Saturday validation was 239 and this morning 130.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
sk..
Master Cruncher http://s17.rimg.info/ccb5d62bd3e856cc0d1df9b0ee2f7f6a.gif Joined: Mar 22, 2007 Post Count: 2324 Status: Offline Project Badges: 
|
I would speculate that someone *** either increased their clock speed, stopped crunching GPU work, turned off HT, or ran out of other tasks (it is slightly faster to crunch one task than 2 or 4).
** looks like HT is on, GPU is being used, and someone decided to run an AV scan or defrag all at once! Or possible started using the battery. |
||
|
|
yose-ue
Cruncher Joined: Dec 27, 2008 Post Count: 21 Status: Offline Project Badges:
|
I have had two results returned as valid a147 and a162 and one returned as invalid a169. The valid results ran uninterrupted. I suspended a169 after its first checkpoint and ran childhood cancer for a short time (not long enough for it to checkpoint) and restarted a169 it ran uninterrupted until completion. The box (leave application in memory while suspended) is not checked so memory was cleared.
|
||
|
|
HutchNYC
Advanced Cruncher United States Joined: Nov 27, 2005 Post Count: 97 Status: Offline Project Badges:
|
Any word on whether or not the B.SE units will be released today?
---------------------------------------- |
||
|
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges: 
|
Few notes on the recent beta. We have encountered a 10% validation error rate (invalids). This is higher than is desired. We have sent some data to the researchers to help us with our validation rules. Until these rules are updated we will not be starting the B.se type. Which at this point may not happen today.
Thanks -Uplinger |
||
|
|
HutchNYC
Advanced Cruncher United States Joined: Nov 27, 2005 Post Count: 97 Status: Offline Project Badges:
|
Thanks for the update and quick response!
---------------------------------------- |
||
|
|
Randzo
Senior Cruncher Slovakia Joined: Jan 10, 2008 Post Count: 339 Status: Offline Project Badges: 
|
Thanks a lot uplinger.
Great job. Please keep us informed. |
||
|
|
|