| 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: strange points claimed and points granted |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 17
|
|
| Author |
|
|
kateiacy
Veteran Cruncher USA Joined: Jan 23, 2010 Post Count: 1027 Status: Offline Project Badges: 
|
Here are results from the last 3 HFCC WUs returned by one of my machines:
----------------------------------------HFCC_ L4_ 00513239_ L4_ 0001_ 0-- kate-jetway64 Valid 2/16/11 04:49:00 2/17/11 12:50:02 24.42 1,622.6 / 138.8 HFCC_ L4_ 00537057_ L4_ 0000_ 0-- kate-jetway64 Valid 2/16/11 13:04:38 2/17/11 03:58:55 10.47 58.0 / 1,334.0 HFCC_ L4_ 00511305_ L4_ 0000_ 0-- kate-jetway64 Valid 2/15/11 22:48:35 2/17/11 01:22:23 24.08 133.2 / 108.8 This is a dual-core Atom D525, so the third WU on the list looks normal -- 100-odd points for a 24-hr run. The status logs look normal for all 3 WUs. What could have caused the huge grant for the the 2nd WU, and the huge overclaim for the first?  Edit to add more info: This machine is running 3 HFCC WUs and 1 GPUgrid WU. It runs BOINC 6.10.17 under Ubuntu 10.04. It is slow but reliable -- regularly gets repair jobs.  [Edit 1 times, last edit by kateiacy at Feb 17, 2011 6:15:31 PM] |
||
|
|
ashrader330
Advanced Cruncher Joined: Jan 6, 2008 Post Count: 97 Status: Offline Project Badges:
|
I have a couple of these very weird points awarded work units also. They all seem to have run normally with no errors. No wing men for either. These two ran on a core i7 920.
---------------------------------------- Not that I don't appreciate the point bump but wow!HFCC_ L4_ 00509848_ L4_ 0000_ 0-- ashrader-desktop Valid 2/15/11 17:36:46 2/17/11 12:52:00 5.49 88.1 / 1,457.1 HFCC_ L4_ 00518985_ L4_ 0000_ 0-- ashrader-desktop Valid 2/16/11 04:41:57 2/17/11 12:52:00 6.21 99.7 / 1,697.5  Run time: 4.2y HPF2, 6.9y FAAH, 7.9y HFCC, 20.8y HCC, 26.0y CEP2, 26.0y MCM, 2.1y UGM, 2.0y OET WU: 4.8k HPF2, 12.3k FAAH, 12.4k HFCC, 135k HCC, 34.3k CEP2, 43.8k MCM, 4.2k UGM, 19.7k OET |
||
|
|
anhhai
Veteran Cruncher Joined: Mar 22, 2005 Post Count: 839 Status: Offline Project Badges: 
|
every once in a while, I get those strange credit granted too. I also have noticed a few strange request, but I believe it happens to all of us. Many people have asked about this, but the only answer we got were that there is some sort of random number generator at work in the backgrounds.
----------------------------------------I believe Sekerob says that there will be a new credit granting system that will be implemented sometime soon (should be done by 2100, possibly even 2050  ) that would resolve these strange occurrences. |
||
|
|
kateiacy
Veteran Cruncher USA Joined: Jan 23, 2010 Post Count: 1027 Status: Offline Project Badges:
|
Thanks for the responses, ashrader330 and anhhai. Glad to know it happens to others and isn't just somethng wierd that my machine has started doing.
----------------------------------------I wonder whether it's just coincidence that ashrader330's odd WUs reported in within 2 minutes of when one of mine did. |
||
|
|
anhhai
Veteran Cruncher Joined: Mar 22, 2005 Post Count: 839 Status: Offline Project Badges:
|
here is what happening on one of my machines
----------------------------------------X0000063460851200601231416_ 1-- Pending Validation 2/17/11 15:44:02 2/18/11 01:18:03 1.35 63.3 / 0.0 X0000063180025200602152113_ 2-- Pending Validation 2/17/11 15:28:47 2/18/11 01:01:40 1.39 65.3 / 0.0 X0000063180722200602011556_ 2-- Pending Validation 2/17/11 15:28:10 2/18/11 00:59:19 1.45 67.9 / 0.0 HFCC_ L4_ 00568038_ L4_ 0001_ 0-- Valid 2/17/11 12:18:06 2/17/11 23:40:29 4.01 67.4 / 471.9 X0000063450120200602031818_ 0-- Valid 2/17/11 13:46:49 2/17/11 23:33:36 1.41 23.8 / 21.3 X0000063451234200601271531_ 1-- Valid 2/17/11 13:40:59 2/17/11 23:18:39 1.44 24.4 / 27.7 HFCC_ L4_ 00560332_ L4_ 0000_ 1-- Valid 2/17/11 11:47:56 2/17/11 23:03:18 3.55 60.9 / 42.2 X0000063440434200602171834_ 1-- Pending Validation 2/17/11 12:33:33 2/17/11 22:01:54 1.37 23.4 / 0.0 X0000063440401200602171836_ 0-- Valid 2/17/11 12:33:13 2/17/11 21:56:22 1.37 23.4 / 23.4 the granted on that one in bold is absurd. However, if you notice the ones on top are asking for over 40 creidt/hr, while the lower ones are asking for only about 17 credits per hour. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
 Project wide some fuse shorted as the impact was over 2 credit per hour extra on 60.3 CPU years. Since points is the base of calculating computational power, there was some 10 TFL extra for HFCC on Thursday... but I'm seeing more credit fest in the fringes, and my resistance is non-existent, futile anyhow :P |
||
|
|
kateiacy
Veteran Cruncher USA Joined: Jan 23, 2010 Post Count: 1027 Status: Offline Project Badges:
|
Sekerob, by "project-wide" you mean all of HFCC, right, not all of WCG? I'm running only HFCC right now, and I'm curious as to whether this is happening elsewhere.
----------------------------------------So far, I've gotten a total of 3 WUs on 2 different hosts with whopping big points granted. And it seems as if, once a machine has received a "bonus," it overclaims for subsequent WUs. I didn't think the hosts got feedback about the WUs they submitted. Hmm, maybe our crunching machines are turning into Watson and learning from experience..  |
||
|
|
TimAndHedy
Senior Cruncher Joined: Jan 27, 2009 Post Count: 267 Status: Offline Project Badges:
|
Perhaps something is going on, my Linux Hex Core is never granted its requested number, on the 17th I see one where I was granted more than requested.
----------------------------------------HFCC_ L4_ 00565296_ L4_ 0000_ 1-- HexComputer Valid 2/17/11 07:23:15 2/17/11 17:22:03 4.06 141.1 / 141.1 HFCC_ L4_ 00569529_ L4_ 0001_ 0-- HexComputer Valid 2/17/11 06:48:23 2/17/11 15:49:02 3.09 98.7 / 113.5 I grabbed all my approx 3 hour units for the last few days on this system. The 65's seem more like I remember. Are they messing with the scoring system for linux? HFCC_ L4_ 00573440_ L4_ 0001_ 0-- HexComputer Valid 2/17/11 17:22:03 2/18/11 01:24:35 3.08 117.8 / 99.8 HFCC_ L4_ 00569529_ L4_ 0001_ 0-- HexComputer Valid 2/17/11 06:48:23 2/17/11 15:49:02 3.09 98.7 / 113.5 HFCC_ L4_ 00542331_ L4_ 0001_ 0-- HexComputer Valid 2/16/11 17:40:07 2/17/11 04:11:27 3.06 96.1 / 70.4 HFCC_ L4_ 00533895_ L4_ 0001_ 0-- HexComputer Valid 2/16/11 15:20:27 2/17/11 00:22:01 3.05 95.9 / 85.1 HFCC_ L4_ 00535279_ L4_ 0000_ 0-- HexComputer Valid 2/16/11 14:24:09 2/16/11 23:34:35 2.90 91.0 / 87.9 HFCC_ L4_ 00526178_ L4_ 0000_ 0-- HexComputer Valid 2/16/11 08:07:51 2/16/11 19:15:11 3.02 95.1 / 65.7 HFCC_ L4_ 00524879_ L4_ 0001_ 0-- HexComputer Valid 2/16/11 07:15:04 2/16/11 17:40:06 2.85 88.2 / 66.4 The numbers look more in line with my Windows system. Which I consider to be approx equivalent in processing power per core. HFCC_ L4_ 00577630_ L4_ 0000_ 0-- QuadComputer Valid 2/17/11 16:00:54 2/17/11 23:52:31 2.87 79.3 / 98.0 HFCC_ L4_ 00555320_ L4_ 0000_ 0-- QuadComputer Valid 2/17/11 01:10:30 2/17/11 09:42:56 2.98 81.9 / 105.2 HFCC_ L4_ 00503684_ L4_ 0000_ 1-- QuadComputer Valid 2/16/11 21:50:53 2/17/11 06:51:49 2.87 80.0 / 79.7 HFCC_ L4_ 00535166_ L4_ 0001_ 0-- QuadComputer Valid 2/16/11 15:20:04 2/17/11 00:37:54 2.87 79.7 / 99.0 HFCC_ L4_ 00526922_ L4_ 0001_ 0-- QuadComputer Valid 2/16/11 08:55:36 2/16/11 18:04:25 3.13 87.2 / 102.2 Update: Just had this odd one come through HFCC_ L4_ 00576547_ L4_ 0000_ 0-- HexComputer Valid 2/17/11 15:49:03 2/18/11 04:11:02 4.26 158.5 / 444.0 [Edit 1 times, last edit by TimAndHedy at Feb 18, 2011 6:13:54 AM] |
||
|
|
TimAndHedy
Senior Cruncher Joined: Jan 27, 2009 Post Count: 267 Status: Offline Project Badges:
|
Sekerob, by "project-wide" you mean all of HFCC, right, not all of WCG? I'm running only HFCC right now, and I'm curious as to whether this is happening elsewhere. I checked my two linux systems running HCC and I don't see anything that looks like a difference. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Better I'd written "science-wide", and in past have proposed to the admin to change all references to a sub-project to be called systematically *Research* or *Science* to avoid the confusion.
----------------------------------------What this was kicking it off I don't know, but few days ago, impacting c4cw/hcmd2, the claims were suddenly 2-3 credit per hour higher on my Linux quad, though there had not been a benchmark. Then it takes about 20 results to sort itself out again... gradually gravitating back to the actual claim. Really stopped wondering, but it does irritate when then pushing the client into overclaiming "outlier" territory and then credit being given per the consistent low-lier/underclaimer. See that allot on DDDT2 off late. So, stopped begging for fixes, and take the "profit" when it falls my way. In this case, widespread, it turns the whole TeraFLops/Petaflops performance discussion into a pass-time divorced from any reality... yes 10TFL extra was recorded for HFCC per yesterday. It's unlikely to get corrected... take the money and run, without flinching ;>) edit: 2-3 credit "per hour" inserted. [Edit 1 times, last edit by Former Member at Feb 18, 2011 9:44:55 AM] |
||
|
|
|