1) Message boards : Number crunching : Last wu crunched - kind of a sad feeling (Message 5012)
Posted 8 Nov 2004 by Granite T. Rock
Post:
You might find some WU's trickle out as deadlines expire and an additional WU needs to be resent. While my one computer has been out of work for a few days... Another one of my computers received 3 new units November 7th.
2) Message boards : Number crunching : @Markku - homogeneous redundancy (Message 4526)
Posted 28 Oct 2004 by Granite T. Rock
Post:

> The use of homogeneous redundancy should allow computational errors to be
> eliminated, leaving only the numerical variation to worry about. Once the
> range of legitimate variation is known then subsequent analysis can cope
> easily with it.


ahhhhhhhhhhhh... (Lightbulb turning on moment)... Thanks... I like it when that happens
3) Message boards : Number crunching : @Markku - homogeneous redundancy (Message 4501)
Posted 27 Oct 2004 by Granite T. Rock
Post:
This may be a silly question:

If there is not cross-platform consistency would LHC not want to eliminate this inconsistancy by finding out the source of it rather than ensuring that the same work unit is processed by the same platform?

My reasoning is that if two seperate platforms produce distinct and different results, then how I know which result or even if the results are valid? I could however, see how homogeneous redundancy being used as an aid in tracking down and understanding how the different platforms behave differently and use that information to make the sourcecode stronger in the future.

Thanks!
4) Message boards : Number crunching : results getting shmucked when returned (Message 4494)
Posted 27 Oct 2004 by Granite T. Rock
Post:
I noticed a smilar trend as well. However on my stats it seems to be slowly updating the 0's with appropriate numbers. When the outcome columns says unkown, I am guessing that means although it has received the result, the server hasn't had a chance to actually look inside and start the validation process. Once the Outcome colunmn changes, you should notice that the other fields are updated. Also remembering, as you implied, that it is normal to have some (but not all) 0 second results.

Happy Crunching
5) Questions and Answers : Windows : Too many 0 second reports (Message 4374)
Posted 26 Oct 2004 by Granite T. Rock
Post:
The problem seemed to clear itself up. I noticed that many of the 0 second ones also said "new" beside them even though downloaded. MAny of them are not having a time show up beside them. My guess is it was the servers way of saying I've received them but haven't even had a chance to see how long they were.
6) Questions and Answers : Windows : Too many 0 second reports (Message 4323)
Posted 25 Oct 2004 by Granite T. Rock
Post:
Hi,

I know that 0 second WU's are normal and to be expected. However, all but 1 of my work units list themselves as being 0 seconds. This is contrary to what I have observed on my computer where several LHC WU's have made it up to the 40 minute mark before being returned. As well, some of the work units listed under my account are 0 while other people who processed the same data were in the thousands of seconds. I am unsure if this is something to be worried about or not. I'm not so much worried about the credit as making sure my WU's are being processed correctly and are useful.

I'm on windows XP Sp1, 512 megs of ram, Boinc 4.13.

http://lhcathome.cern.ch/results.php?userid=5549

Thanks!



©2024 CERN