I'm running a "Do You Feel Lucky? (GFN World Record)" search. Two questions points after running two of these numbers:
I've noted that when these numbers are allocated for testing to me , the completion target date is significantly under forecasted. In my situation by about an additional 2 months against a estimate of about 25 days. All my other Prime/BOINC calculation estimates are accurate, only the World record numbers are not correct. ( Properties for current number; Issued 3Dec19, for a target of 24Dec19, and an actual CPU estimate achieved at 25Feb20, genefer_extreme_28201252, 31,458,443 GFLOPs)
Secondly, my computation has reached the target CPU effort after 32days of CPU time, but still only shows 70.767% done, (its stuck there it seems), a CPU time of 5hours 11 minutes, and zero estimated elapsed time remaining all showing in the properties screen. The running time continues to increment.
I'm reluctant to try any more of these extreme number types if the process is problematic.
Questions: Is the completion time target correctly computed for these extreme large numbers, and more importantly should I leave my task running or abort after 32 days of CPU effort over about 83 days of elapsed computation? I'm reluctant to abort after all this effort if there is a chance of completion. Any suggestions?