| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: BOINC Agent Support
Thread: Something going on with WU ex379_2A ?? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 22
|
|
| Author |
|
|
depriens
Senior Cruncher The Netherlands Joined: Jul 29, 2005 Post Count: 350 Status: Offline Project Badges: 
|
A lot of CPU time is wasted on this workunit. 11.56 hours from my side.  Shouldn't there be a way that a workunit gets aborted when there are more than 3 errors? At this moment additional workunits are sent out to others and I guess there will probably be more errors.How is the policy of handling these workunits? Will there be new workunits until there is a valid quorum?   |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
No, five errors and the work unit will be automatically rejected and reported to WCG for analysis.
----------------------------------------Correction: four errors. [Edit 1 times, last edit by Former Member at Jun 12, 2006 11:39:13 AM] |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Asked the very question in another thread about 4 weeks ago, no response.....now we know its 5 errors on 1 WU.
----------------------------------------thx @!DePriens.....had quite a few EX....., which crunched all flawlessly. A few took 7 to 9 hours, rest 2 to 4 hours, so must be a fluke..... It's hot in Holland is it not...MBM5 for readings, Speedfan for realtime reading and active automated fan control at specified temp! It's taken 5 minutes CPU time in the 10 days since last boot. 
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I think you did get a response eventually - that's how I know now.
|
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Ah sorry , i did not see that added comment:
----------------------------------------Added: I just reread your question. I think that I did not read it correctly. knreed has said that 4 Errors causes a work unit to be pulled for review but I do not know what the other limits are. Obviously, it will make a difference if every result is different, if 1 is different and another was not returned, etc. The simple title 'Inconclusive' does not give us enough information to determine just which case is happening. The general rule is that the valid solution is determined by a majority of results (minimum size of majority = 3). So 3 identical results plus 2 different results will result in validation. Here you actually are saying, that pulling takes place at 4 errors, or were you meaning 4 inconclusives....... In the DePriens Case, those 3 other crunches are unstoppable and who knows do they turn valid! Enjoy the heat up in the northern zones....still only 18c down south.
WCG
----------------------------------------Please help to make the Forums an enjoyable experience for All! [Edit 1 times, last edit by Sekerob at Jun 12, 2006 10:24:53 AM] |
||
|
|
depriens
Senior Cruncher The Netherlands Joined: Jul 29, 2005 Post Count: 350 Status: Offline Project Badges:
|
Okay, thanks for the info. So if there'll be another error this WU will be rejected.
----------------------------------------The error occured just when the benchmarks started to run: 2006-06-11 21:23:39 [---] Suspending computation - running CPU benchmarks I've had this once more: 2006-05-17 17:10:06 [---] Suspending computation - running CPU benchmarks Temperature is not a problem, yes it's over 30degC over here at the moment, but my processor runs only a few degrees hotter now. 53degC constantly. Normally i'm around 46-47 degC. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
I've had errors coming at me until i unticked the option to keep the suspended process in memory....it writes it to disc and you loose a few minutes as it returns to last checkpoint saved. Also, i changed the 'save to disc' from 60 seconds to 5 minutes a few days ago, which also improved thruput a few percentage points. Not booted in 10 days, so considered the hi reliability.
----------------------------------------Interesting, your whetstone is 100 better, your dhrystone 150 less....always running truly flatline 60c on CPU and 42c on MB and 43c on HD! I'm told this is best for durability.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello depriens,
According to your first post, the 4th Error was returned at 06/12/2006 08:13:17. This should have marked that work unit for review. Another work unit was sent out at 06/12/2006 08:14:09. Apparently the replacement work unit is dispatched before the error count is checked against the limit of 4.  Well, that is something to add to the list of desired program fixes. Lawrence |
||
|
|
depriens
Senior Cruncher The Netherlands Joined: Jul 29, 2005 Post Count: 350 Status: Offline Project Badges:
|
Glad to hear that the error (and this post) was good for something in the end. Hopefully you can fix this minor flaw in a later stage. Thanks for all replies.
---------------------------------------- |
||
|
|
depriens
Senior Cruncher The Netherlands Joined: Jul 29, 2005 Post Count: 350 Status: Offline Project Badges:
|
Another one with the same error as the two errors mentioned above: ex621_02
----------------------------------------And -again- with running the CPU Benchmarks. Only this time on another computer. Strange that this ONLY happens to computers which run BOINC 5.4.9 (2 computers). All my other computers (11) run the previous BOINC version (5.3?) and they never had any problem with running the benchmarks. I'm starting to believe that it has something to do with the newest BOINC client. Have there been any changes to the Benchmark routine? |
||
|
|
|