Joined: 19 Jul 99
I have one computer (ID = 2001659) that cannot download sufficient GPU workunits (WUs) to run continuously. It is a machine with a new CPU and mobo that I turned on after a hiatus of several years. Usage is at 100% now. Yesterday, Sunday 9/11/11, it downloaded WUs in dribs and drabs of 1, 2, 3, 6, 16, at most, at a time. It processes an average of 163 WUs per day, and all I could download yesterday was perhaps 208 of 358 processed. And that was with constant fussing with it: Restarting downloads and forcing updates. It is not that the Seti@home Web site did not have the WUs; at all times it had between 250,000 and 300,000 of them in inventory, ready to hand out. Yet Boinc kept saying, "No workunts sent," with no explanation of why. That machine does have three error WUs in the last few days, none of which were its fault; all had too many triplets to fit in the table, and other users who processed the same WUs experienced the same situation. How could that be an error? If one plots active users v. date (fron 5/16 to 9/11/11) and fits a trend line to the data the regression equation is y = -0.0128x + 532.15, R² = 0.8267, where y is Percent Active Users and x is Date, and this equation explains ~83% of the variance in active users. I other words, if Seti@Home does not want to lose more active users, please consider keeping the servers working and handing out work. If Seti@Home needs more equipment, then please consider a fund-raising drive.
|ID: 1151445 ·|
Joined: 9 Apr 02
I have one computer (ID = 2001659) that cannot download sufficient GPU workunits (WUs) to run continuously.
Because of a rash of "shorties" that were being sent out, work was being downloaded, completed and returned faster than the servers can keep up. This created a sort of traffic jam in the internet pipe used to distribute work and accept returned results.
Also note that even though the servers might say they have 250,000 workunits available, the splitters only know about 100 workunits at a time. So at any given moment, when your computer asks for work, there may not be any in this short queue of 100 until it is filled with available work to be sent.
I other words, if Seti@Home does not want to lose more active users, please consider keeping the servers working and handing out work. If Seti@Home needs more equipment, then please consider a fund-raising drive.
Please note that SETI@Home has never promised to have work available at all times. Suggesting that SETI@Home needs to keep the servers "working and handing out work" does not fit with the Distributed Computing directive of doing big science on a little budget.
While I'm sure the Project Admins would be happy with newer hardware, throwing more money at the problem to meet expectations of promises never made doesn't seem to be a real solution.
|ID: 1151447 ·|
©2016 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.