Message boards :
Number crunching :
New rescheduler
Message board moderation
Previous · 1 · 2 · 3 · 4 · 5 . . . 11 · Next
Author | Message |
---|---|
hiamps Send message Joined: 23 May 99 Posts: 4292 Credit: 72,971,319 RAC: 0 |
Anyone know what I screwed up? 13 July 2010 - 06:22:25 BoincRescheduler V: 0.7 13 July 2010 - 06:27:15 Found: CPU: 705, VLAR: 705, VHAR: 0 13 July 2010 - 06:27:15 Found: GPU: 3519, VLAR: 0, VHAR: 0 13 July 2010 - 06:27:15 No rescheduling needed 13 July 2010 - 06:27:53 Shutting down BOINC client 13 July 2010 - 06:27:57 Shutdown of BOINC client completed 13 July 2010 - 06:28:41 Invalid rsc_fpops_bound < 500000000000000000.000000, total: 4222 13 July 2010 - 06:28:41 Found: CPU: 705, VLAR: 705, VHAR: 0 13 July 2010 - 06:28:41 Found: GPU: 3517, VLAR: 0, VHAR: 0 13 July 2010 - 06:28:41 Rescheduling CPU version: 603 ,Gpu version: 608 planclass: cuda 13 July 2010 - 06:29:32 Copied rescheduled client_state.xml 13 July 2010 - 06:29:32 ERROR: The BOINC client is running, but it shouldn't, so something is wrong 13 July 2010 - 06:29:32 ERROR: Move completed with an error. Official Abuser of Boinc Buttons... And no good credit hound! |
S@NL - eFMer - efmer.com/boinc Send message Joined: 7 Jun 99 Posts: 512 Credit: 148,746,305 RAC: 0 |
Anyone know what I screwed up? That's one of these checks I made and never expected to see. The client is running after it reported it had stopped. Maybe the client did not properly shut down and restarted itself. Maybe the client is a bit sluggish and slow because your tasks are above the safe limit of BOINC. But the state file is copied, so at least shut down the client and restart it to see if it reports any problems. TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking. |
hiamps Send message Joined: 23 May 99 Posts: 4292 Credit: 72,971,319 RAC: 0 |
I did a restart of Boinc and the manager didn't show any error messages. I hope it made the flop change...So now there are 3 reschedulers... Official Abuser of Boinc Buttons... And no good credit hound! |
hiamps Send message Joined: 23 May 99 Posts: 4292 Credit: 72,971,319 RAC: 0 |
Anyone know what I screwed up? My client is fine as far as sluggishness, not sure what you mean by "Safe limit of Boinc" My manager works fine with way more work than this. Something changed around February and my client hasn't had sluggish problems since, and from what I hear none of the top hosts have this problem anymore. People have a whole lot more work than I have. The funny thing is most of my work downloaded last night and it was almost all on the CPU until I rescheduled this morning. I moved 2000 to GPU. Official Abuser of Boinc Buttons... And no good credit hound! |
S@NL - eFMer - efmer.com/boinc Send message Joined: 7 Jun 99 Posts: 512 Credit: 148,746,305 RAC: 0 |
The more WU you have on your system the more overhead. You have a bit more that you need for 3 days. TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking. |
hiamps Send message Joined: 23 May 99 Posts: 4292 Credit: 72,971,319 RAC: 0 |
I am about to go back to lurking mode until this is straitened out. Can't keep up with all the suggestions is so many different threads. Too bad someone can't make a good sticky that others can't add to but would have to be edited by only one person. The stickies get way to sidetracked...As long as I don't have 100's of downloads stuck, my manager can handle twice this amount of work without any communication problems. Official Abuser of Boinc Buttons... And no good credit hound! |
hiamps Send message Joined: 23 May 99 Posts: 4292 Credit: 72,971,319 RAC: 0 |
I am still in Eternal High Priority Mode but have some of those Aug 4 units waiting to run. Will see if they send me back to 100 hours when another goes thru. It did last night just like I thought they would. Official Abuser of Boinc Buttons... And no good credit hound! |
perryjay Send message Joined: 20 Aug 02 Posts: 3377 Credit: 20,676,751 RAC: 0 |
Bet it won't get there. 7/13/2010 11:56:12 AM SETI@home Scheduler request failed: Couldn't connect to server Looks like we are down for the count until Friday. PROUD MEMBER OF Team Starfire World BOINC |
S@NL - eFMer - efmer.com/boinc Send message Joined: 7 Jun 99 Posts: 512 Credit: 148,746,305 RAC: 0 |
V 0.8 Add: Settings tab: Option, start at login. Change: When your try starting the program for the second time, it shows the dialog of the running copy. Fixed: Expert setting "Limit rsc_fpops_bound" sometimes failed to update a workunit. TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking. |
[DPC] hansR Send message Joined: 14 Jul 00 Posts: 47 Credit: 235,829,569 RAC: 8 |
Tested the "Limit rsc_fpops_bound" aetting. Now works OK. Thanks! |
S@NL - eFMer - efmer.com/boinc Send message Joined: 7 Jun 99 Posts: 512 Credit: 148,746,305 RAC: 0 |
V 0.9 Changed: Now updates rsc_fpops_bound on running tasks. Changed: When moving from CPU <-> GPU, the est and bound values will be the calculated ones not the server values. Add: Move work from GPU to CPU. Add: Expert tab: Read config button, will read the config.xml again. Add: Config.xml SETI settings: <est_ratio_cpu_min> <est_ratio_cpu_max> <est_ratio_gpu_min> <est_ratio_gpu_max> These can be used to limit the range of the CPU and GPU ratio on server and calculated work. This is an example of the config.xml file I now use. <config> <seti> <dcf_min>0.8</dcf_min> <dcf_max>1.2</dcf_max> <est_ratio_cpu_min>4</est_ratio_cpu_min> <est_ratio_cpu_max>4</est_ratio_cpu_max> <est_ratio_gpu_min>5</est_ratio_gpu_min> <est_ratio_gpu_max>5</est_ratio_gpu_max> </seti> </config> I use this to get the DCF as close as possible to 1.0. Instead of the 0.1 I had before. Now: Duration correction factor 0.9964010000 The DCF is allowed to go from 0.8 to 1.2. A higher ratio value means shorter runtimes = faster computer. The min and max values are the same, this in effect sets this value. TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking. |
[DPC] hansR Send message Joined: 14 Jul 00 Posts: 47 Credit: 235,829,569 RAC: 8 |
Problem on windows XP running as service. After restart all CPU jobs erroring out. CUDA just starts running. No info in BOINC log ??? (stdoutdae.txt/old) |
S@NL - eFMer - efmer.com/boinc Send message Joined: 7 Jun 99 Posts: 512 Credit: 148,746,305 RAC: 0 |
Running as a service? That's not the preferred mode with a GPU for some time. And I haven't any computer running BOINC as a service, so I can't test it. You not by any chance, have a copy of the BOINC folder before this you could send me. What CPU and GPU versions are you running? And errors of this kind will show up in the regular logging. TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking. |
[DPC] hansR Send message Joined: 14 Jul 00 Posts: 47 Credit: 235,829,569 RAC: 8 |
I know, but running as a service under XP is my preferred configuration for all my windows XP systems. Only 1 system has a CUDA GPU and this works perfect. Problem is only with Vista en W7. No copy of the BOINC folder because it just happened unexpected. Realy nothing in the logging. Maybe boinc was still running as a service and no real restart of boinc or second copy? BOINC manager logged something in its message window about missing ddl's or so, but this didn't show up in the logfiles. Have no more info available. Lost all my CPU jobs, and not getting new jobs :-( Example of one of the jobs: http://setiathome.berkeley.edu/result.php?resultid=1661259358 |
[DPC] hansR Send message Joined: 14 Jul 00 Posts: 47 Credit: 235,829,569 RAC: 8 |
Have changed BOINC installation to not run as service. The outage is coming soon and this systeem needs a lot of jobs ;-) |
S@NL - eFMer - efmer.com/boinc Send message Joined: 7 Jun 99 Posts: 512 Credit: 148,746,305 RAC: 0 |
Have changed BOINC installation to not run as service. The outage is coming soon and this systeem needs a lot of jobs ;-) You could try moving work over from the GPU to the CPU. Start with a small number, this should get you GPU work as well. TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking. |
hiamps Send message Joined: 23 May 99 Posts: 4292 Credit: 72,971,319 RAC: 0 |
Looks as though the reschedulers are still needed, have over 100 VLars on GPU's so far. Official Abuser of Boinc Buttons... And no good credit hound! |
Hellsheep Send message Joined: 12 Sep 08 Posts: 428 Credit: 784,780 RAC: 0 |
I believe the reason running as a service under Win 7 or Vista doesn't work for GPU crunching is due to the way the drivers run under those OS's, something to do with WDDM. I think Joe posted a bit about this, i'll have a look and see if i can find the thread. Also, the vlar change has only taken effect on Beta as far as i know. If all goes well at Beta, it will be released here soon. - Jarryd |
hiamps Send message Joined: 23 May 99 Posts: 4292 Credit: 72,971,319 RAC: 0 |
I believe the reason running as a service under Win 7 or Vista doesn't work for GPU crunching is due to the way the drivers run under those OS's, something to do with WDDM. I think Joe posted a bit about this, i'll have a look and see if i can find the thread. Cool wasn't sure, downloads are working this AM. Had to set to no new work...YAY Official Abuser of Boinc Buttons... And no good credit hound! |
S@NL - eFMer - efmer.com/boinc Send message Joined: 7 Jun 99 Posts: 512 Credit: 148,746,305 RAC: 0 |
V 1.0 Add: A balloon text when the BOINC client, is not running for 3 minutes. Changed: Don't use the mutex to check if the BOINC client is running. Changed: Check if the BOINC client is running just before the copy of the new client state file. Fixed: the est_ratio_xpu min and max values where reversed. TThrottle Control your temperatures. BoincTasks The best way to view BOINC. Anza Borrego Desert hiking. |
©2024 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.