I guess you got a new datafile and the new one has shorter WUs. You know, you always download datafiles which are good for at least a couple dozen WUs, which are usually of roughly the same size. Once all WUs from that datafile are crunched, you download a new one, which can of course have differently sized WUs. Hope this helps.
I kind of wish I could get more of the short duration work units. Did they introduce an extra-long work unit recently? Because of the long time to crunch these work units, and because I run other projects, I only have three results from the past week. Crunch times were roughly 12.5 hours, 15 hours, and 1 day 3 hours. Why did I end up with something that's taking double the previous units? The one work unit currently in my queue is estimating 23.5 hours. I'm pretty sure these crunch times are indepenent of what other projects I may be running, so it looks like I'm getting longer work units than before. Can anyone tell me why this is happening? I looked around the board a bit, but didn't really see anything. A pointer is fine, if the answer is here somewhere. Thanks.
I kind of wish I could get more of the short duration work units. Did they introduce an extra-long work unit recently? Because of the long time to crunch these work units, and because I run other projects, I only have three results from the past week. Crunch times were roughly 12.5 hours, 15 hours, and 1 day 3 hours. Why did I end up with something that's taking double the previous units? The one work unit currently in my queue is estimating 23.5 hours. I'm pretty sure these crunch times are indepenent of what other projects I may be running, so it looks like I'm getting longer work units than before. Can anyone tell me why this is happening? I looked around the board a bit, but didn't really see anything. A pointer is fine, if the answer is here somewhere. Thanks.
You had the same problem as (s)he did. Looking in the middle of your long-running WU, you'll find:
2006-10-05 02:14:38.6710 [normal]: Start of BOINC application 'projects/einstein.phys.uwm.edu/einstein_S5R1_4.24_windows_intelx86.exe'.
2006-10-05 02:14:38.6867 [normal]: Started search at lalDebugLevel = 0
2006-10-05 02:14:42.7023 [normal]: Found checkpoint-file 'Fstat.out.ckp'
Failed to read checkpoint-counters from 'Fstat.out.ckp'!
2006-10-05 02:14:42.7023 [normal]: No usable checkpoint found, starting from beginning.
Some of the ideas we batted around there may help you too.
Strange. Why this problem should have happened now and never before, I don't know. It looks like the main suggestion in the thread you mentioned is to leave applications in memory. I guess I'll try that and hope it works. If I have more problems or questions, I'll post them in that thread. Thanks for the info.
Small Work Units!
)
Work units come in 2 sizes, long and short.
Short ones on my machine take about 55 minutes. Long ones about 10 hours.
Kathryn :o)
Einstein@Home Moderator
Hi Dave, I guess you got a
)
Hi Dave,
I guess you got a new datafile and the new one has shorter WUs. You know, you always download datafiles which are good for at least a couple dozen WUs, which are usually of roughly the same size. Once all WUs from that datafile are crunched, you download a new one, which can of course have differently sized WUs. Hope this helps.
I kind of wish I could get
)
I kind of wish I could get more of the short duration work units. Did they introduce an extra-long work unit recently? Because of the long time to crunch these work units, and because I run other projects, I only have three results from the past week. Crunch times were roughly 12.5 hours, 15 hours, and 1 day 3 hours. Why did I end up with something that's taking double the previous units? The one work unit currently in my queue is estimating 23.5 hours. I'm pretty sure these crunch times are indepenent of what other projects I may be running, so it looks like I'm getting longer work units than before. Can anyone tell me why this is happening? I looked around the board a bit, but didn't really see anything. A pointer is fine, if the answer is here somewhere. Thanks.
RE: I kind of wish I could
)
Have a look at this thread by Moorkens.
You had the same problem as (s)he did. Looking in the middle of your long-running WU, you'll find:
2006-10-05 02:14:38.6710 [normal]: Start of BOINC application 'projects/einstein.phys.uwm.edu/einstein_S5R1_4.24_windows_intelx86.exe'.
2006-10-05 02:14:38.6867 [normal]: Started search at lalDebugLevel = 0
2006-10-05 02:14:42.7023 [normal]: Found checkpoint-file 'Fstat.out.ckp'
Failed to read checkpoint-counters from 'Fstat.out.ckp'!
2006-10-05 02:14:42.7023 [normal]: No usable checkpoint found, starting from beginning.
Some of the ideas we batted around there may help you too.
Thanks everyone!
)
Thanks everyone!
RE: Have a look at this
)
Strange. Why this problem should have happened now and never before, I don't know. It looks like the main suggestion in the thread you mentioned is to leave applications in memory. I guess I'll try that and hope it works. If I have more problems or questions, I'll post them in that thread. Thanks for the info.