Message boards :
Number crunching :
SETI@home v8.22/23 Windows GPU applications support thread
Message board moderation
Author | Message |
---|---|
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
Please post issues and questions regarding Windows OpenCL apps here. SETI apps news We're not gonna fight them. We're gonna transcend them. |
TheTruthIsOutThere Send message Joined: 12 Feb 15 Posts: 29 Credit: 265,079 RAC: 0 |
So something strange start occuring yesterday and I didn't really realise it till now what is generally up so heres the story I have 4 cpu cores and 2 gpus on my laptop running on boinc and both gpus work on seti@home Yesterday I got a massive bunch of postponed tasks from seti@home and then it went to waiting to run. A bunch disappeared (or maybe a issue with the SoG ones caused the other seti tasks i normally do have an issue but then worked itself out) but however I don't see any errors on task list on seti page. However it has never done the whole waiting to run on seti@home tasks before and I couldn't figure out why Today I looked and I noticed from my screenshot from lastnight and from today (after a reboot/shutdown last night) that its the 8.22 opencl_ati5_SoG_nocal tasks only doing this before anyone asks yes all my boinc folders are excluded/ignored from my scanner wondering if this maybe a bug or something? i'lll include a bunch of screen shots from last night to today. if there is anyway to make the logs show me more information u need let me know and i'll do so and post. Computer ID 8166299 -- tasks show no errors after dec 21 (my past problem i solved) I did suspend all my tasks and tried to leave the opencl_ati5_SoG_nocal tasks unsuspended bu they kept going from postpond to waiting to run and back and forth (the two on my machine atm) Ideas? |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
https://setiathome.berkeley.edu/forum_thread.php?id=80754&postid=1838601 SETI apps news We're not gonna fight them. We're gonna transcend them. |
TheTruthIsOutThere Send message Joined: 12 Feb 15 Posts: 29 Credit: 265,079 RAC: 0 |
thanks looked at it and restarted boinc as well and those two didnt get aborted and i guess the other ones from last night did get aborted. will try manual abort and see if it all goes alright |
TheTruthIsOutThere Send message Joined: 12 Feb 15 Posts: 29 Credit: 265,079 RAC: 0 |
hey me again sorry to be yanking your chain but I thought all was well until my client started to receive some new tasks. I noticed this in the log 2016-12-31 6:00:35 PM | SETI@home | [error] No app version found for app setiathome_v8 platform windows_intelx86 ver 823 class opencl_ati5_SoG_nocal; discarding 11ja09af.28756.90200.6.33.84_0 I do have an app_config.xml file but it strictly is for max_concurrent tasks. and nothing more. any ideas? I did check in client_state.xml and didn't see anything regard 8.23 just 8.22 maybe my boinc didn't get the updated files for 8.23? if so a way to push it? cheers n happy new years |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
I think you have to fetch the master scheduler list to force it to get the proper 8.23 ATI_SOG app. The 8.22 tasks were abandoned by the project because of an error. Try suspending network communication and then force an Update. Let it time out with the error message and then turn network communication back on. Then Suspend the project, wait a bit and then turn the project back on. Hopefully when you look in the logfile you will see the "fetching master scheduler list" message and then right after that you should see the download of the 8.23 app. I'm not sure if this is exactly the correct order of steps but I have done something similar and caused my system to fetch the master scheduler list. Maybe someone else more knowledgeable can point to the exact steps. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
marmot Send message Joined: 15 May 99 Posts: 144 Credit: 1,220,664 RAC: 0 |
I think you have to fetch the master scheduler list to force it to get the proper 8.23 ATI_SOG app. The 8.22 tasks were abandoned by the project because of an error. Try suspending network communication and then force an Update. Let it time out with the error message and then turn network communication back on. Then Suspend the project, wait a bit and then turn the project back on. Hopefully when you look in the logfile you will see the "fetching master scheduler list" message and then right after that you should see the download of the 8.23 app. I'm not sure if this is exactly the correct order of steps but I have done something similar and caused my system to fetch the master scheduler list. Maybe someone else more knowledgeable can point to the exact steps. Just clicking the project and choosing "Reset Project" while in advanced options. BOINC will refresh all the files and leave your app_config.xml alone. I haven't used the simple interface in years so not sure how to do a project reset from that. |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
Yes, that will certainly accomplish the desired act. But unless you have crunched all your tasks, that also dumps all your current work because "resends" haven't been turned on in a long time. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
marmot Send message Joined: 15 May 99 Posts: 144 Credit: 1,220,664 RAC: 0 |
Shouldn't be a concern in this case since all my machines only started getting the error once the older WU's were finished and they were out of work. |
TheTruthIsOutThere Send message Joined: 12 Feb 15 Posts: 29 Credit: 265,079 RAC: 0 |
tried it but didn't work. ill keep trying other ideas |
TheTruthIsOutThere Send message Joined: 12 Feb 15 Posts: 29 Credit: 265,079 RAC: 0 |
tried the reset as well as didn't work. |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
tried the reset as well as didn't work. Reset should have worked as it sends all necessary files again for a "blank slate" condition. Only thing I can think of is the server being in a funky state of not having all the ATI 8.23 changes migrated through the database or it doesn't recognize your card as an ATI or something. Going to take some of the experts to decipher what is going on probably. Maybe post over in the BOINC Questions and Problems forum. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
marmot Send message Joined: 15 May 99 Posts: 144 Credit: 1,220,664 RAC: 0 |
tried the reset as well as didn't work. It's rare but possible a file was locked or corrupted and the project wasn't properly refreshed. I'd try doing the steps of manual reset by saving your app_config.xml somewhere, remove the SETI project using BOINC manager, use file manager to make sure the SETI project folder is deleted, then re-add the project and copy back in your app_config. If that doesn't work then... (their bad WU's mentioned in the link above should be purged by now) I'd scan the hard drive system for errors. |
TheTruthIsOutThere Send message Joined: 12 Feb 15 Posts: 29 Credit: 265,079 RAC: 0 |
I will def try this and let uknow if it works I set to not accept new tasks as I run my queue out first. then once that's done by noon tmr ill do it. thx for the idea |
Garry Send message Joined: 7 Jul 02 Posts: 40 Credit: 535,102 RAC: 1 |
Two things. No complaints, just reports, in case they help. 1) I have an inexpensive Win10 computer (read: low capability). I give it all the BOINC it can handle, in addition to my uses for it. Typically, it's running a CPU-only Rosetta task and a GPU S@H task and they cooperate well. Today, the GPU task (for the 8.22ati application) competed very unsuccessfully for CPU while at 97% complete. In the BOINC Manager, both the elapsed time and the remaining time were increasing (though the remaining time was increasing perhaps a tenth as fast or less than the elapsed time); the percent complete stayed nearly constant. After it had done that for ten hours or so, I paused the Rosetta task. No surprise: Suddenly the S@H task had lots of CPU to play with and promptly completed its work. I now have another S@H task (8.22 ati). 2) When I heard 8.23 had come out, I didn't expect to see any more 8.22 tasks. I'm still only getting 8.22 ati and 8.22 ati5 (presumably computers with more capabilities get tasks from a wider set of applications). I hope that helps. |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
Well then, obviously this isn't a case of one single computer having issues. The changeover to 8.23 ATI tasks isn't migrating through the servers for some reason. Probably needs to be brought to Eric's attention. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
marmot Send message Joined: 15 May 99 Posts: 144 Credit: 1,220,664 RAC: 0 |
Checked my 4 machines running SETI and both ATI and nVidia are still getting 8.22. No errors tho. |
TheTruthIsOutThere Send message Joined: 12 Feb 15 Posts: 29 Credit: 265,079 RAC: 0 |
ya still get 8.22 after a delete everything and re load in the project. and stil same error about no app version for 8.23 so I think it is something server side possibly. resuming basic work again on what I can |
TheTruthIsOutThere Send message Joined: 12 Feb 15 Posts: 29 Credit: 265,079 RAC: 0 |
also 8.22 tasks work as far as I knew it was only 8.23 for the SoG ones |
Raistmer Send message Joined: 16 Jun 01 Posts: 6325 Credit: 106,370,077 RAC: 121 |
https://setiathome.berkeley.edu/apps.php Windows/x86 8.20 (opencl_intel_gpu_sah) 14 Dec 2016, 0:46:29 UTC 25,365 GigaFLOPS SETI apps news We're not gonna fight them. We're gonna transcend them. |
©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.