zero credit for work completed on time?
28 Apr 2005 12:38:26 UTC
Topic 189062
(moderation:
Not sure what was going on here, the result page shows
28 Apr 2005 9:36:17 UTC Over Success Done 28,230.56 69.72 0.00
The work unit was completed well before the deadlines (1st of May), the other two computer gets the credit due at 88.02
zero credit for work completed on time?
)
> Not sure what was going on here, the result page shows
>
> 28 Apr 2005 9:36:17 UTC Over Success Done 28,230.56 69.72 0.00
>
> The work unit was completed well before the deadlines (1st of May), the other
> two computer gets the credit due at 88.02
>
===============================================================================
The unit:
Result ID 3601902
Name H1_0811.9__0812.0_0.1_T15_Run2_1
Workunit 895451
Created 24 Apr 2005 21:30:49 UTC
Sent 24 Apr 2005 22:02:27 UTC
Received 28 Apr 2005 9:36:17 UTC
Server state Over
Outcome Success
Client state Done
Exit status 0 (0x0)
The reason why no credits are granted:
Validate state Invalid
Claimed credit 69.7174112941022
Granted credit 0
application version 4.78
The posible cause:
Software Conflict..?
Time to open this box and remove the dust..?
CPU Running to hot..?
Defective ram module..?
Power failure..?
HD corupted, failure, Crash..?
etc...
Greetz
"The FUTURE is only a PARTICLE away from the PRESENT and the PAST."
There have been many problems
)
There have been many problems in the past with the validation proces in which a non-Windows computer (Linux, Mac-OS) is "computing" against a WinXP machine; The results from such machines differs, causing the validator to declare the result INVALID.
The validator has improved recently, up to now my results (linux and MacOSX) are all validated, however it could still be possible that some are just unlucky.
In your results I see a lot of starting/stopping, this could have been contributed to the invalid result; it was at least suspected then.
There are numerous threads about this to read on if interested you could do a search at the beginning of the message board.
John,
Mac's AMD's and Intel's have
)
Mac's AMD's and Intel's have different floating point processors.
The AMD and Intel differ only in the last bit or so on one or two floating point calculations.
Looking at your pc's results, it seems to be a rare event that you get an invalid workunit. So, at the current rate of error, I'd say it's not that big a deal.
I think that for the math that this project is doing, that the differences between a Mac, Intel or AMD floating point processors should be irrellevent.
I doubt that the sensor data from the Ligo is measured to 64 bit resolution.
Maybe there's a math error in the calculations that they're using.
> I doubt that the sensor
)
> I doubt that the sensor data from the Ligo is measured to 64 bit resolution.
The resolution of the data is not a likely source of error. However, with most models the minor errors and differences in the FPUs can lead to vastly different outcomes. Iterative processes with errors even at the far end of significant digits can slowly "migrate" up to where they alter the significant digits near the decimal point.
> Maybe there's a math error in the calculations that they're using.
It is far more likely that it is an error as I have described above.
> Not sure what was going on
)
> Not sure what was going on here, the result page shows
The result status is quite clear if you return a result too late, as indicated here: resultid=3618171.
To summarise:
Validate state Result was reported too late to validate
Claimed credit 77.4942187788677
Granted credit 0
Sometimes it happens, too bad for me....but that's life.