Ok thanks, i'll chalk it up then to a project which stresses hardware more than the others mentioned and/or has more stringent verification.
I don't know about Rosetta, but we are definitely more picky than SETI concerning the results. And we are strictly CPU-bound, which puts more load on the CPU (actually FPU) itself than e.g. the memory, which might also be different to other projects.
I'm sorry to say that after more than 150 days of trouble-free project operation, one of our five validator instances went on the rampage at about 10:30 UTC 16-11-2006. I learned about this from the message boards and from an email (thanks Timothy!) around 00:15 UTC 17-11-2006. I restarted the errant validator and all is well again.
My apologies for this glitch - I don't know what went wrong on the server side but experience shows that in spite of our best efforts to run a stable project, these things do sometimes happen.
Credit hounds (and others): I am granting credit for ALL those results which were marked as invalid or validate error during the time window above.
[Subsequent edit/addition]: another validator instance started to show the same problem the next day. I have restarted ALL the validator instances, and granted credit for the second round of problematic results as well.
OK, I have restarted all five instances of the validator: one of the OTHER instances was showing problems also. I hope this fixes it. I will modify the credit-granting script to give credit for these results also.
...Credit hounds (and others): I am granting credit for ALL those results which were marked as invalid or validate error during the time window above...
Actually you had ALREADY been granted credit for the result, but the web page 'workunit.php' was not displaying this properly. If you had clicked on the result in question then you would have seen the granted credit on that next web page that would have been displayed.
Anyway I have modified the 'workunit.php' web page so that it now correctly shows the granted credit.
Validate errors
)
I don't know about Rosetta, but we are definitely more picky than SETI concerning the results. And we are strictly CPU-bound, which puts more load on the CPU (actually FPU) itself than e.g. the memory, which might also be different to other projects.
BM
BM
Dear Einstein@Home
)
Dear Einstein@Home participants,
I'm sorry to say that after more than 150 days of trouble-free project operation, one of our five validator instances went on the rampage at about 10:30 UTC 16-11-2006. I learned about this from the message boards and from an email (thanks Timothy!) around 00:15 UTC 17-11-2006. I restarted the errant validator and all is well again.
My apologies for this glitch - I don't know what went wrong on the server side but experience shows that in spite of our best efforts to run a stable project, these things do sometimes happen.
Credit hounds (and others): I am granting credit for ALL those results which were marked as invalid or validate error during the time window above.
[Subsequent edit/addition]: another validator instance started to show the same problem the next day. I have restarted ALL the validator instances, and granted credit for the second round of problematic results as well.
I apologize again,
Bruce Allen
OK, I have restarted all five
)
OK, I have restarted all five instances of the validator: one of the OTHER instances was showing problems also. I hope this fixes it. I will modify the credit-granting script to give credit for these results also.
Bruce
RE: Would the credit show
)
Actually you had ALREADY been granted credit for the result, but the web page 'workunit.php' was not displaying this properly. If you had clicked on the result in question then you would have seen the granted credit on that next web page that would have been displayed.
Anyway I have modified the 'workunit.php' web page so that it now correctly shows the granted credit.
Cheers,
Bruce