Message boards :
Number crunching :
new tasks?
Message board moderation
Author | Message |
---|---|
Miklos M. Send message Joined: 5 May 99 Posts: 955 Credit: 136,115,648 RAC: 73 |
I do not seem to be getting more than just a few new tasks. What am I doing wrong? Thanks |
Bill G Send message Joined: 1 Jun 01 Posts: 1282 Credit: 187,688,550 RAC: 182 |
My suggestion would be to reload the drivers from the NVidia site. I suspect that you recently had an OS update and your drivers were changed by that. After reloading the drivers restart BOINC/Seti and hopefully all will be working.. If this does not word start the first 30 lines or so of the log after Seti Starts. SETI@home classic workunits 4,019 SETI@home classic CPU time 34,348 hours |
Miklos M. Send message Joined: 5 May 99 Posts: 955 Credit: 136,115,648 RAC: 73 |
Thank you. |
Miklos M. Send message Joined: 5 May 99 Posts: 955 Credit: 136,115,648 RAC: 73 |
The method of reinstalling the gpu drivers almost proved a disaster for me. My computer always went into the sleep mode instantly, so I could not do anything, yet later when I checked it it was set for sleep after 15 minutes. |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
The method of reinstalling the gpu drivers almost proved a disaster for me. My computer always went into the sleep mode instantly, so I could not do anything, yet later when I checked it it was set for sleep after 15 minutes. If Nvidia drivers are suspect, it is always best to completely remove them before reinstalling. On Windows, the DDU uninstaller is the best tool. Then download the drivers directly from Nvidia for reinstallation. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
rob smith Send message Joined: 7 Mar 03 Posts: 22200 Credit: 416,307,556 RAC: 380 |
Try disabling sleep while you install the drivers - they can be "somewhat fussy" if sleep tries to do things during the installation. Also, do a "clean" installation - this is an option under the heading of "advanced installation". That said, if you are using "sleep" it is quite probable that your computer is getting as many tasks as BOINC thinks you require to fill your cache. Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
Miklos M. Send message Joined: 5 May 99 Posts: 955 Credit: 136,115,648 RAC: 73 |
I downloaded the gpu from NVIDIA saved it and then deleted the old NVDIA programs. Why would a setting of "sleep after 15 minutes of not use" confuse BOINC? Puzzled on that one. |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
If your computer spends the majority of its time in suspend/sleep, then the APR's for both cpu and gpu tasks will drop dramatically. The APR rate is how BOINC determines how much work to send you to be completed by the task deadlines. Set sched_op_debug in the Log options to see how many seconds of work for both cpu and gpu are requested on the scheduled server connection. It sounds like Seti is sending you exactly what it thinks can be handled by your system. To get more work, allow the system to stay running longer. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
Miklos M. Send message Joined: 5 May 99 Posts: 955 Credit: 136,115,648 RAC: 73 |
Since I certainly do not have computer skills, all I can see is that I am getting plenty of work for PrimeGrid and since I started Collatz (since SETI does not send any work this way) I also get lots of work there. This seems to be a SETI issue on my end, but I am not good enough to follow the directions sent so far, sorry. I have been a cruncher for almost 20 years, I would love to keep crunching. Best |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13736 Credit: 208,696,464 RAC: 304 |
Your Resource share settings determine which projects get work, and which ones don't. The more projects you run, the smaller your cache should be- to avoid some projects being starved of processing time for extended periods while meeting the Resource share settings for other projects with large caches & short deadlines. Grant Darwin NT |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
You have overcommitted debt to your other projects relative to Seti which has extremely long deadlines compared to other projects. You need to increase the share percentage of Seti relative to your other projects. If you use the BOINC Manager and under either Tools or Options, you will find Event Log options. Set the sched_op_debug toggle in the Options and then hit Apply. Open the Event Log and read the entries in the Event Log at the next 5 minute scheduled server connection to Seti. That Event Log option will show you the amount of cpu and gpu seconds the Seti scheduler is going to send you based on your ability to process it and the amount of cache you've set within the normal 100cpu/100gpu task limit. This is an example of one my hosts scheduled connection requests. You should see something similar with that option in the Event Log if you have toggled it on correctly. It will also show you the requested amount of work for each of your other projects. You seem to be requesting much more work for your other projects than for Seti. Sat 03 Nov 2018 04:19:11 PM PDT | SETI@home | Sending scheduler request: To fetch work. Sat 03 Nov 2018 04:19:11 PM PDT | SETI@home | Reporting 12 completed tasks Sat 03 Nov 2018 04:19:11 PM PDT | SETI@home | Requesting new tasks for CPU and NVIDIA GPU Sat 03 Nov 2018 04:19:11 PM PDT | SETI@home | [sched_op] CPU work request: 1216974.99 seconds; 0.00 devices Sat 03 Nov 2018 04:19:11 PM PDT | SETI@home | [sched_op] NVIDIA GPU work request: 111060.93 seconds; 0.00 devices Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
Miklos M. Send message Joined: 5 May 99 Posts: 955 Credit: 136,115,648 RAC: 73 |
Thank you both. What settings would you recommend? I actually run three projects once I receive work from SETI. |
Miklos M. Send message Joined: 5 May 99 Posts: 955 Credit: 136,115,648 RAC: 73 |
I would welcome any easy to follow suggestion on how to allocate the computer usage amounts. Thanks |
rob smith Send message Joined: 7 Mar 03 Posts: 22200 Credit: 416,307,556 RAC: 380 |
I would go for: Cache: "store at least" = 1 day "store additional" = 0.01day Resource shares: Favourite project = 1000 Next favourite = 100 Least favourite = 10 or if you want the least favourite to be a last resort backup project set it to 0 - then it will only get work when there is no work available from the other two. It will take some time (weeks) for things to settle down, but settle they will. Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
Tom M Send message Joined: 28 Nov 02 Posts: 5124 Credit: 276,046,078 RAC: 462 |
I would welcome any easy to follow suggestion on how to allocate the computer usage amounts. I would use the website(s) to set the resource # for the relative amount that you want to devote to each project. If you wanted a Seti dominated system you could use something like 100 for Seti and 1 for the others. After you have made those changes at each project website you would hit the "update" button on BOINC Manager for each project under the projects tab to have those settings show up on your system. I would use the BOINC Manager local configuration to set the CPU cores usage to the nearest percentage that will give you a whole number of cores for "90%". I have forgotten to look at your system. This advice presumes at least 8 threads on the cpu. If you are running a 4 core non-hyper threading cpu, the advice might be 75%. For 2 cores, I am not sure I would do anything. Make sure all the other "local configuration" items are the same as you have on your seti website. Hit save. And off you go. HTH, Tom A proud member of the OFA (Old Farts Association). |
Miklos M. Send message Joined: 5 May 99 Posts: 955 Credit: 136,115,648 RAC: 73 |
I guess I may have made the correct changes in the three projects. Perhaps someone could look at it and tell me? |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
We have no knowledge of what your share resources for each project are. That is visible only to you. I would say you still don't have enough resource share for Seti since I see no work at all on your hosts. Seti isn't asking for any work because BOINC is allocating your hosts capabilities primarily to your other projects like PrimeGrid. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
To give you and idea of how I've allocated resources. Seti - 900 Einstein - 50 GPUGrid - 50 Total =1000 All the projects get crunch time eventually. No problem with Einstein completing work on time. GPUGrid gets into priority mode a lot of the time because of its shorter deadlines. And I never get the 50% bonus for completing work within 24 hours. Except occasionally on Tuesday's on very long outrages. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
rob smith Send message Joined: 7 Mar 03 Posts: 22200 Credit: 416,307,556 RAC: 380 |
Keith, did you know that the total for all resoursse share can be more tthan 1000 as each project can have a 'score' of 1000? Bob Smith Member of Seti PIPPS (Pluto is a Planet Protest Society) Somewhere in the (un)known Universe? |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
Yes, I knew. I just do 2 or 3 projects besides Seti and the 1000 total made it easier for easy granularity with the lesser projects. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
©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.