| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Outsmart Ebola Together
Thread: Collapsing points claims on OET1 |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 47
|
|
| Author |
|
|
TimAndHedy
Senior Cruncher Joined: Jan 27, 2009 Post Count: 267 Status: Offline Project Badges: 
|
Somewhat related, whats up with this.
This unit ran for 13 hours but my wingman claimed 40.5? Very odd. That would bring the points down quickly. Workunit Status Project Name: Outsmart Ebola Together Created: 02/11/2015 13:14:10 Name: OET1_0000333_xMBGP-OM_rig_7910 Minimum Quorum: 2 Replication: 2 Result Name App Version Number Status Sent Time Time Due / Return Time CPU Time / Elapsed Time (hours) Claimed/ Granted BOINC Credit OET1_ 0000333_ xMBGP-OM_ rig_ 7910_ 2-- 719 Valid 2/11/15 16:01:09 2/12/15 14:10:31 16.81 595.2 / 317.8 OET1_ 0000333_ xMBGP-OM_ rig_ 7910_ 1-- 719 Error 2/11/15 15:57:44 2/11/15 16:00:56 0.00 46.3 / 0.0 OET1_ 0000333_ xMBGP-OM_ rig_ 7910_ 0-- 719 Valid 2/11/15 15:57:40 2/19/15 14:55:51 13.38 40.5 / 317.8 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Your wingman got hit with the outlier safety rule, yours just not. The 40 could have been the average for the short units on that machine, then when it got to do a long one, some multiplier got checked and the machine was tied to the whipping post.
Does the 595 claimer really run 'normal' 37+/hr grant? S.N.A.S.U. |
||
|
|
TimAndHedy
Senior Cruncher Joined: Jan 27, 2009 Post Count: 267 Status: Offline Project Badges:
|
Does the 595 claimer really run 'normal' 37+/hr grant? . That does look to be on the high side. More like 20. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
It's of course Banana Republic the way the credits are being arrived at, the average just happening by sheer luck-factor [don't know how that got programmed in], the ~19/hour then working out not having been dealt short.
The whole average OET credit seems fairly high anyhow with just under 27/hour, with WCG 7 day mean running 25.8/hour, another top-hat excuse not to fix it, yet  |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7844 Status: Offline Project Badges: 
|
I do not want to beat this point issue to death, but I did try to take a look at the points just for the intellectual exercise. I have mainly only had my Core2Duo at 2.66 ghz on Linux running OET1. I used a sample of 597 units from Feb 10 to Feb 19 to see if I could spot any trends. I did but just barely as the relationships are pretty tenuous. Those units which took less than a 15 minutes (1/4 hour) gave the greatest points per hour value on average, but the range was very wide.
----------------------------------------For< 15 Minutes Minimum Points per Hour 15 Maximum Points per Hour 189 Average Points per Hour 122 Efficiency 96.45% For 15 to 60 Minutes Minimum Points per Hour 18 Maximum Points per Hour 153 Average Points per Hour 63 Efficiency 98.99% For > 60 Minutes Minimum Points per Hour 16 Maximum Points per Hour 71 Average Points per Hour 44 Efficiency 99.02% The shortest unit was 140 seconds and the longest was 9 hours, 10 minutes, 48 seconds. (The longest unit got 16 points per hour). About the only trend I can see is the longer the unit, on average, the more efficiently it runs. The longest unit had an efficiency of 98.97%, but I know this unit had one computer shutdown about midway through. Also on this machine I was granted an average of -2.44 fewer points than claimed, although the range on this was very wide varying from +94 to -80. I am curious to see if it settles into a more stable pattern over time. I do note the overall average for the project since restarting is very stable at about 26 Points per hour. Don't read too much into this as this still represents just one machine over 10 days and the project is still very young. Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
TimAndHedy
Senior Cruncher Joined: Jan 27, 2009 Post Count: 267 Status: Offline Project Badges:
|
Picked up a few repair units
My wingman is making the lowest claim I have seen. 62 hours for 52 points. It will be interesting to see what my system claims. Workunit Status Project Name: Outsmart Ebola Together Created: 02/11/2015 20:33:35 Name: OET1_0000337_xZAGP-OM_rig_10473 Minimum Quorum: 2 Replication: 2 Result Name App Version Number Status Sent Time Time Due / Return Time CPU Time / Elapsed Time (hours) Claimed/ Granted BOINC Credit OET1_ 0000337_ xZAGP-OM_ rig_ 10473_ 2-- - In Progress 2/22/15 03:44:50 2/25/15 15:44:50 0.00 0.0 / 0.0 OET1_ 0000337_ xZAGP-OM_ rig_ 10473_ 1-- 719 Pending Validation 2/12/15 03:44:54 2/15/15 20:47:47 62.05 51.6 / 0.0 OET1_ 0000337_ xZAGP-OM_ rig_ 10473_ 0-- 719 Error 2/12/15 03:44:37 2/22/15 03:44:54 0.00 50.3 / 0.0 |
||
|
|
KWSN-A Shrubbery
Senior Cruncher Joined: Jan 8, 2006 Post Count: 476 Status: Offline Project Badges: 
|
52 points is the outlier code kicking in.
---------------------------------------- |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
More off this stuff underway... a whole set is racing through at under 8 minutes of batch 349-353, whilst a few of 331 are about to hit 24 hours at 56-66%, copy _4 as in, more than some are not waiting on the outlier news :(
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
OMG, got 248 queued on the octo, with 13 minutes TTC, as told by the server, which is quick to respond to this fast flow... no wonder there was a high server load this morning.
 |
||
|
|
Speedy51
Veteran Cruncher New Zealand Joined: Nov 4, 2005 Post Count: 1326 Status: Offline Project Badges: 
|
OMG, got 248 queued on the octo, with 13 minutes TTC, as told by the server, which is quick to respond to this fast flow... no wonder there was a high server load this morning. Far out that is a lot. What are your runtimes like?  |
||
|
|
|