| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Computing for Clean Water Forum
Thread: A bunch of Inconclusives... |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 20
|
|
| Author |
|
|
Bugg
Senior Cruncher USA Joined: Nov 19, 2006 Post Count: 271 Status: Offline Project Badges: 
|
In the past 24 hours I've returned 11 units that have a state of Inconclusive. Should I just assume that this can be a normal thing now and then, or that something else has happened (and if so, what)?
----------------------------------------I haven't changed a thing on my machine in weeks now. So I'm unable to imagine how it could be something on my end.  i5-12600K (3.7GHz), 32GB DDR5, Win11 64bit Home |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7576 Status: Offline Project Badges: 
|
Do not worry about them. They will resolve in good time. The system is merely checking your machine.
----------------------------------------Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
Bugg
Senior Cruncher USA Joined: Nov 19, 2006 Post Count: 271 Status: Offline Project Badges:
|
So units that are Inconclusive are sent out purposely to check our machines accuracy? Are they known already to have issues or not give a good/correct result?
----------------------------------------i5-12600K (3.7GHz), 32GB DDR5, Win11 64bit Home |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello Bugg,
3 ways to get an inconclusive: 1) a new unrated computer. 2) an ERROR (perhaps because of a bad batch) makes rerating necessary. 3) randomly selected computers are checked to catch systematic problems. A bunch of Inconclusives on several computers are almost certainly caused by a previous batch of bad work units. Perhaps the algorithm fails over a range or perhaps the work units were made incorrectly. In any case, just let it pass. A single computer that persistently needs rerating should be watched suspiciously for overclocking problems, etc. Lawrence |
||
|
|
Bugg
Senior Cruncher USA Joined: Nov 19, 2006 Post Count: 271 Status: Offline Project Badges:
|
Ok thanks Lawrence. This is not a new computer, but it is overclocked, and has been running great for a few weeks now at this new speed. I'll keep an eye over the next few days to see if I get too many more.
----------------------------------------Is it common that once overclocked for a few weeks, it would start to show problems like this, suddenly? Just trying to narrow it down. :) i5-12600K (3.7GHz), 32GB DDR5, Win11 64bit Home |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Shrug. Just keep an eye out for future Invalids or Errors and examine them carefully. Each one might lead to a number of Inconclusives. If it happens again, I would consider reducing the overclocking.
Lawrence |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
So units that are Inconclusive are sent out purposely to check our machines accuracy? Are they known already to have issues or not give a good/correct result? Bugg, I might be reading your questions wrongly, but I thought I would try to prevent any possible misconception about units that get a result status of Inconclusive. The English language and its usage can so easily lead to ambiguity! In case you meant, "Are [units that are Inconclusive] known already to have issues ...", no, there's nothing special about such units. They are normal research units that need crunching. It's the subsequent server processing of results that gives Inconclusive or not, as others have implied.If you haven't already, look at the FAQ Work Unit progress state descriptions for the definition of Inconclusive, but also scroll right down to the bottom of that page where there is a more detailed explanation of Credit Method for Zero Redundancy distribution projects. If your machine is behaving, your Inconclusive results should turn Valid once the 2nd copy has been returned and shown to match yours. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The "Inconclusive" was looked at some years ago [off-line], and one suggestion made was to name them "Pending Revalidation"... less worrying.
Under the old server, the series of Inconclusives to force a re-assessment of a device, [and all devices are subjected to this], could be as much as 15. Don't know how many are needed under the server 700 reliability rating logic. If there are more, then start looking at a device problem, OCing certainly one of those sources. If it is frequent reoccurring, look again. In the long run, count on 10% being checked for Zero Redundant science distributions (FAAH/HFCC/C4CW)... intermittent a few or in a series. Can't remember to have seen these on CEP2. --//-- |
||
|
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 2977 Status: Offline Project Badges: 
|
Sek, just to confirm, yes occasionally CEP2 units are sent for "revalidation" (i.e. set as "inconclusive") on occasion. I know - as I've just received one earlier on this morning...
---------------------------------------- |
||
|
|
Bugg
Senior Cruncher USA Joined: Nov 19, 2006 Post Count: 271 Status: Offline Project Badges:
|
Ok, thank you for the much clearer explanation. It makes a lot more sense now. :)
----------------------------------------i5-12600K (3.7GHz), 32GB DDR5, Win11 64bit Home |
||
|
|
|