Col323 2012-10-08 13:15:23 | I had done a clean install on a new laptop and submitted points from it yesterday. I run this laptop offline in a command line and used manualsend.bat to submit results. I run it concurrently with World Community Grid in Boinc, letting Muon1 soak up spare cycles as it can. Prior to submitting about 1.3 MB of work from the laptop, I checked my stats via Free-DC to see I had about 44,000 points for the day, which is about my usual pace. I was quite surprised to see after my submission from the laptop, I had -25,316 points for the day. http://stats.free-dc.org/stats.php?page=user&proj=dpad&name=%5BEVGA%5DCol323 The points are no great loss, but I'm more concerned about continuing to run DPAD on this laptop. I don't want to keep going backwards! If it helps, here is my config.txt. I configured it with config_nonet.bat, but I have set the number of threads by hand while I was playing with its interaction with the Boinc client, set the priority to Background, and turned off updates of the lattice files. This last setting was causing the client to hang during the server issues a few weeks ago. I made all these changes in Notepad. Thanks for the help! Resolution (XxY, 'auto' or 'span'): auto Priority for background running ([N]ormal, [L]ow, [B]ackground): B Threads (number or 'auto'): 4 Auto-send results ([Y]es, [N]o): N Results upload method ('FTP' or 'HTTP'): HTTP Use proxy server for HTTP downloads ('auto', 'no' or server,port): auto Auto-save interval (seconds; 0 for no save): 240 Update lattice files from web every N hours (0=don't): 0 Download sample results file after a number of hours (0=don't, min. 6): 0 Sample file behaviour - use [L]atest only, or [A]ccumulate: L Sample file URL (.bin allowed): http://stephenbrooks.org/muon1/samplefiles/{lattice}_100.bin Results file maximum size before pruning (MB, 0 to disable): 50 Keep mode ([B]est results, [F]irst, [L]atest or [R]andom): B Keep proportion (%): 50 Removed results store (or 'none'): results_removed.dat Sleep on battery power: Y CPU throttle (%): 100 Preferred lattices (comma-separated list or 'none'): none Particles per extra thread (limits threading overhead): 100 Rechecks for best-so-far results (min. 5): 5 FTP command call: ftp -script.ftp Use passive (PASV) mode for FTP transfers: N Use old FTP calling method (as pre-v4.34; can hang, but more compatible with ZoneAlarm): N TrialType ratios: Random=0;Mutate=3;Crossover=3;Interpolate=3;Extrapolate=3;MuSpherical=3;MuOne=3;Extreme=1;LocalGrad=3;TopoSmooth=1;CellDouble=1;CellDelete=1;CellCommute=1; Also, here are yesterday's entries in my sendlog: 20121007-234929 394 results sent to 81.0.198.134 20121007-234944 391 results sent to 81.0.198.134 20121007-235008 392 results sent to 81.0.198.134 20121007-235020 142 results sent to 81.0.198.134 |
Stephen Brooks 2012-10-08 17:15:22 | Not quite sure what the recent server problems could have done to various clients, so can you do a clean install but keep your old "results.dat"? Then copy the tail end of results.dat into results.txt and resend the results since your going-backwards problem using manualsend.bat. |
Col323 2012-10-08 19:12:19 | Done. I checked Free-DC before sending, I'm a touch over 41,000 points for the day. We'll see what shakes out. Here is today's sendlog: 20121008-180844 394 results sent to white-panther.mine.nu 20121008-180903 391 results sent to white-panther.mine.nu 20121008-180921 392 results sent to white-panther.mine.nu 20121008-180940 142 results sent to white-panther.mine.nu |
Col323 2012-10-09 13:54:23 | Ok, I'm quite confused. I ended up with 82,287 for the day yesterday, which seems like a good sign. But I thought there were filters in place to keep people from cheating by resubmitting points from their results.dat file. Which, if they were corrupted on the first send, then it does make sense that they passed through the filters yesterday. But I parsed my submitted results and calculated a submission of over 58,000 points. And I'm not the brightest bulb, but 41,000+58,000 > 82,287. So I'm still concerned about this laptop. I may pause my other submissions for a day and then upload results generated from the laptop's new installation and see if the numbers balance. |
Stephen Brooks 2012-10-09 18:10:38 | Maybe you submitted more results off the end of results.dat than were actually missing? (So you counted 58000 but in fact fewer were missing). I see activity in your log around 2012-Oct-08 20:05 with various results being rejected, probably from the results.dat upload. Overall, that stats run received 1318 results 58060 Mpts from you, of which 507 results (26407.9 Mpts) were accepted and 811 results (31652.1 Mpts) were rejected as duplicates. Since then, there have been no rejections of what you've uploaded. |
Col323 2012-10-10 16:35:03 | I did two successful experiments today. First I uploaded the laptop's latest 80KB of results. That reconciled exactly. Then, I uploaded some unsent results which were generated prior to the new install. Those also came through exactly as expected. I believe the laptop is all set to crunch! Thank you for your help and verification! |