| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Beta Testing
Forum: Beta Test Support Forum
Thread: Large DSFL Beta Test March 16, 2012 |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 44
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
See Uplinger's post about the validators earlier in this thread (Mar 16, 2012 8:42:37 PM). The validators stay off until the techs have made use of the results in the database. We are past his timeframe of "usually 24 hours after the start" I wonder what they are doing with the unvalidated data for so long? Most unusual. It should also mean there will be a sudden dump of beta work into the system to fix up the inconclusive/errors/etc, instead of the usual long trickle. |
||
|
|
KWSN - A Shrubbery
Master Cruncher Joined: Jan 8, 2006 Post Count: 1585 Status: Offline |
Since they're using this data to size the upcoming work there is no rush. And yes, there will be a rush of repairs/inconclusive when they turn them on.
---------------------------------------- Distributed computing volunteer since September 27, 2000 |
||
|
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges: 
|
Greetings all,
I did not forget about turning the beta validators/assimilators on. I was working through a bug with the program that captures the data for each target/ligand calculation. This is used for sizing the workunits in the future. | target_000010-1.pdbqt | 298.893863636364 | | target_000010-2.pdbqt | 429.520707317073 | | target_000010-3.pdbqt | 615.119326923077 | | target_000010-4.pdbqt | 603.945133684211 | | target_000010-5.pdbqt | 416.281704109589 | | target_000010-6.pdbqt | 591.098853846154 | | target_000010-7.pdbqt | 437.718242857143 | | target_000010-8.pdbqt | 637.222636170213 | | target_000010-9.pdbqt | 281.406623076923 | As you can see from just the first 9 targets listed the average cpu time per ligand varies greatly. I have recently just turned the assimilator/validator back on so there may be some resends going out soon. Thanks, -Uplinger |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I have recently just turned the assimilator/validator back on All DSFL Beta went Valid here. so there may be some resends going out soon. Ooh, yes please.Edit: Ha, as soon as I had sent the above, one appeared. The orignal two were Inconclusive. [Edit 1 times, last edit by Former Member at Mar 20, 2012 6:28:41 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Keith, the first nine is kiddy stuff. At times see checkpoint intervals of 45 minutes, but yes, you'd want to know how many surprises to pack in a WU and not set off a "Monster Ball" thread ;O)
------------------------------------------//-- edit: Them being the production application used to test a new target, I'd not expected anything different than 21 out of 21 valid. \o/ [Edit 1 times, last edit by Former Member at Mar 20, 2012 6:55:38 PM] |
||
|
|
astroWX
Advanced Cruncher USA Joined: Sep 1, 2007 Post Count: 56 Status: Offline Project Badges: 
|
Seven of seven validated.
[Unlike Tony, I haven't lucked-into any resends. Oh, well, seven boxes lurk ...] |
||
|
|
astrolabe.
Senior Cruncher Joined: May 9, 2011 Post Count: 496 Status: Offline |
34 WU with no unexpected errors and 1 still PV due at an AWOL wingman. Another 4 days of Beta on the board.
|
||
|
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges:
|
Sek,
You will still see the large checkpoints still. What this does is help us size the actual work units. How many jobs to put into a work unit. For example, if you have lots of ligands sizes that run about 5 minutes each, then to make an 8 hour work unit you'll need 96 jobs to make an 8 hour work unit. But if the target is say estimating 600 seconds per ligand from the test above, then we will consider each ligand that would take 5 minutes on a normal target to take 10 minutes. Then instead of getting 96 jobs within a workunit, you'll get 48. Which will come out to 47 check points and 1 finish. The monster checkpoints are ligands that are difficult, meaning they are estimated to take 30 minutes each or longer. Then with the difficult target, it could estimate it would take 1 hour per ligand target calculation. It is a two headed beast, meaning the ligands counts for some of the processing time and the target counts for some of the processing time. All of this so that we can keep the workunits within reason and not run for say 4 days :P if we didn't estimate and packaged 96 jobs within each workunit no matter what the estimates were, you could get in theory 96 jobs that run for an hour each. Ahh, the joys :) Thanks, -Uplinger |
||
|
|
nanoprobe
Master Cruncher Classified Joined: Aug 29, 2008 Post Count: 2998 Status: Offline Project Badges: 
|
All but 2 tasks of 6 pages validated. Wingmen are still in progress. I expect them to go no reply/resend.
----------------------------------------Sending them both to me would be a bonus. 
In 1969 I took an oath to defend and protect the U S Constitution against all enemies, both foreign and Domestic. There was no expiration date.
  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
My Athlon 64 recently picked up a No Reply:
BETA_ BETADSFL_ 00040-39_ 0000000_ 0008_ 2-- - In Progress 22/03/12 19:47:06 24/03/12 10:11:06 0.00 0.0 / 0.0 BETA_ BETADSFL_ 00040-39_ 0000000_ 0008_ 1-- 624 Pending Validation 16/03/12 19:11:04 17/03/12 09:20:02 5.72 70.9 / 0.0 BETA_ BETADSFL_ 00040-39_ 0000000_ 0008_ 0-- - No Reply 16/03/12 19:11:03 22/03/12 19:11:03 0.00 0.0 / 0.0 |
||
|
|
|