| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Beta Testing
Forum: Beta Test Support Forum
Thread: what is up with this? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 14
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
these WU's are taking around 15 seconds to finish.
BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724585_ 724585_ 1-- Dell Valid 5/22/09 17:10:35 5/22/09 17:11:29 0.00 0.0 / 0.0 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724540_ 724540_ 1-- Dell Valid 5/22/09 17:09:45 5/22/09 17:10:52 0.00 0.0 / 0.0 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724529_ 724529_ 1-- Dell Valid 5/22/09 17:09:24 5/22/09 17:10:35 0.00 0.0 / 0.0 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724519_ 724519_ 0-- Dell Valid 5/22/09 17:09:07 5/22/09 17:10:14 0.00 0.0 / 0.0 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724504_ 724504_ 2-- Dell Valid 5/22/09 17:08:51 5/22/09 17:09:58 0.00 0.0 / 0.0 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724492_ 724492_ 1-- Dell Valid 5/22/09 17:08:36 5/22/09 17:09:24 0.00 0.0 / 0.0 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724493_ 724493_ 0-- Dell Valid 5/22/09 17:08:36 5/22/09 17:09:45 0.00 0.0 / 0.0 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724473_ 724473_ 0-- Dell Valid 5/22/09 17:08:08 5/22/09 17:09:07 0.00 0.0 / 0.0 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724459_ 724459_ 1-- Dell Valid 5/22/09 17:07:47 5/22/09 17:08:51 0.00 0.0 / 0.0 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724433_ 724433_ 2-- Dell Valid 5/22/09 17:07:25 5/22/09 17:08:36 0.00 0.0 / 0.0 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724408_ 724408_ 0-- Dell Valid 5/22/09 17:06:48 5/22/09 17:07:47 0.00 0.0 / 0.0 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724355_ 724355_ 2-- Dell Valid 5/22/09 17:05:51 5/22/09 17:07:09 0.00 0.0 / 0.0 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724306_ 724306_ 0-- Dell Valid 5/22/09 17:05:19 5/22/09 17:06:48 0.00 0.0 / 0.0 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724289_ 724289_ 1-- Dell Valid 5/22/09 17:04:59 5/22/09 17:06:48 0.01 0.1 / 0.1 BETA_ CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_ 397_ 724264_ 724264_ 0-- Dell Valid 5/22/09 17:04:27 5/22/09 17:05:19 0.00 0.0 / 0.0 |
||
|
|
oliverstirling
Advanced Cruncher United Kingdom Joined: May 7, 2007 Post Count: 107 Status: Offline Project Badges: 
|
I'm in the same boat, have had 30 wu's complete in the space of a few seconds. Think its due to the way the wu's are being split to avoid the monster ones. Knreed explained it in the announcements page:
----------------------------------------Each workunit is assigned a time limit. The beta workunits running now have a time limit of 1 hour. When the time limit is reached, the application checks to see how many positions out of the total required for the workunit have been computed. If at least 60% of the total positions have been computed, then it continues processing. If less than 60% have been computed, then it exits and returns its results to the server. Sekerob posted this example: First copy does 30% in 1 hour or 15 positions Child copy does 50% in 1 hour or 25 positions Closing copy does remaining 20% in 20 minutes, the last 10 positions. Only in this case the closing wu's are only a few seconds long. [Edit 3 times, last edit by oliverstirling at May 22, 2009 6:24:39 PM] |
||
|
|
superempie
Cruncher Joined: Nov 16, 2004 Post Count: 10 Status: Offline Project Badges: 
|
I experience 15-25 sec WU's only on 32 bit machines (XP and Server 2003). On my 64bit machines (both XP64) they take about an hour or so.
|
||
|
|
knreed
Former World Community Grid Tech Joined: Nov 8, 2004 Post Count: 4504 Status: Offline Project Badges: 
|
I'm looking at this now. A bunch of 'micro' workunits are being sent out now.
What appears to be happening is that certain positions are much more difficult to compute then other positions. Our algorithm was based on the assumption that an average position would take a few seconds to compute and the longest would take at most 100 times that. We are seeing single positions take a reasonably powerful processor 75 minutes or so to compute. This means that a single position is returned to the server and the child workunits are chopped up into single workunits as well. Unfortunately, the other positions are proving to be quite easy to compute so you get these super short workunits. We are looking into this now to see what we can do to avoid this. Fortunately, it looks like the longest is around 75 minutes or so. This suggests that full sized workunits should avoid this issue. I'm also opening up the limit for num workunits you can process since I suspect some folks will start hitting their limit shortly. |
||
|
|
knreed
Former World Community Grid Tech Joined: Nov 8, 2004 Post Count: 4504 Status: Offline Project Badges:
|
A quick tutorial on result names:
BETA_CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_397_724355_724355_2 BETA_CMD2_ 0001-GPDAA.clustersOccur-TPM1A.clustersOccur_397 Base workunit name - represents the original workunit we sent out. 724355_724355 If these are found at the end of the name, then this is a child of the original workunit. The numbers indicate which positions are being computed. This one indicates that it is a 1 position workunit. |
||
|
|
knreed
Former World Community Grid Tech Joined: Nov 8, 2004 Post Count: 4504 Status: Offline Project Badges:
|
Ok - we have purged 30,000 of the micro workunits from the beta test.
We have added some controls to limit the number of children created. This will avoid cases where these micro workunits are created. It will in some cases require 'grand children' to be created, but that is more reasonable then the fun we just had. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Ok - we have purged 30,000 of the micro workunits from the beta test. We have added some controls to limit the number of children created. This will avoid cases where these micro workunits are created. It will in some cases require 'grand children' to be created, but that is more reasonable then the fun we just had. Once again, Thanks for the work you do Knreed. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
knreed, I gather that even though you purged some 30,000 units, you didn't purge all. I've gotten 3 micro (about a minute) WUs since your post.
Or are there being children created at microsize still? |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
knreed, I gather that even though you purged some 30,000 units, you didn't purge all. I've gotten 3 micro (about a minute) WUs since your post. Or are there being children created at microsize still? I think the rest of his post explained it, "We have added some controls to limit the number of children created. This will avoid cases where these micro workunits are created. It will in some cases require 'grand children' to be created, but that is more reasonable then the fun we just had." I have also received the minute long Micro WU's. I prefer getting those than the 10 second ones. What if the minute long WU's are actually the "grand children"? |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Well, possible, but very unlikely. The last one lasted for 35s so... yeah. Gotten yet another Beta WU since my last post, I'm feeling rather lucky, took about 35min. So again, I'm feeling uncertain whether those minute-long WUs can be considered grand children.
And aren't there posts about some WUs taking several hours? ... Ah. http://www.worldcommunitygrid.org/forums/wcg/viewthread?thread=25710 I'd say this is one of the grand children, even if it's reported from a slow computer. |
||
|
|
|