PrimeGrid
Please visit donation page to help the project cover running costs for this month

Toggle Menu

Join PrimeGrid

Returning Participants

Community

Leader Boards

Results

Other

drummers-lowrise
1) Message boards : Generalized Cullen/Woodall prime search : Not getting work (Message 115479)
Posted 942 days ago by Profile rebel9
Thanks for trying to help but all wrong. I don't know what you're looking at. As stated, I have only these WUs set in these groups for CPU. The default group has all CU projects enabled, including these two so, even if you were right, I should still get some of this work.

Here's home: https://i.imgur.com/jRZSKm7.png and here's work: https://i.imgur.com/cOxuq2t.png.

Cheers.
2) Message boards : Generalized Cullen/Woodall prime search : Not getting work (Message 115310)
Posted 947 days ago by Profile rebel9
Can anyone help me to understand why I haven't received a single GCW-LLR or GCW-sieve WU when I have had only those projects selected as CPU projects since mid-last year? I have 2 computers, in my home and work groups and the prefs for those groups are set to accept only these projects for CPU work. I'm mystified.

Thanks.
3) Message boards : Generalized Fermat Prime Search : WUs aborted, restarted, not sure...? (Message 90008)
Posted 1765 days ago by Profile rebel9
Hi,

Thanks for taking the time to provide that info.

Why is it being stopped and started so often? That's not normal.

I don't know. Perhaps managing CPU load? Suspending when on battery? Unhelpfully, there's no time scale provided in the log.

Look for "stdoutdae.txt" in Boinc data directory. This is main Boinc log file; if it's enabled, it should contain all clues (why tasks were paused).

I do use a temperature controller. Laptops run far too hot with BOINC otherwise. I killed one in the past.

So why I usually recommend do not crunch on laptops, they're not designed to work at full CPU load for a long time.

What is temperature controller? Is an external application? Could it connect to Boinc and pause tasks when CPU becomes too hot? If you have even a littlest problem with cooling system (a dust in the fan/heatsink or blocked airflow path, degraded thermal paste on the CPU, etc.), CPU temperature could jump above thermal controller limit in just a few seconds. Tasks will be immediately paused, CPU will cool down for a while, task resumed... then immediately paused again, and so on, without any crunching progress.


I've just seen the new, extended genefer prefs. Which options would you suggest are best in my circumstances? I might try some of the 4-day WUs and see how that works out.

Start with limiting number of tasks (CPU cores used by Boinc) to 1. This'll decrease load.

Check temperature of the CPU (SpeedFan is popular very small freeware program).

I suggest do not run LLR or Genefer on your hardware. If you want to help Primegrid, run sieves projects instead.

Anyway, if you want to try GFN, then:

GFN-15 is the smallest and uses slow transform which keeps CPU cooler. Good for testing.

The only optimal project for CPU crunching is GFN-18. It still uses fast transform, but it will end soon (I expect no more then a one month). Average time per workunit, even on CPUs without AVX, is about 5 hours (AVX and AVX2 will be much faster). All other GFN projects are already using slow transform now, so I see no sense to run them on CPU anymore (GPU is much, much faster).


1) 0 bytes, unfortunately, so no help.

2) I agree. Yes, yes & yes. I use tthrottle, which is excellent. I've been running on many computers, including several laptops, for over a decade now. I do have to be careful on laptops. That was a hard lesson to learn. ;-) I run tthrotttle on every machine as a precaution against temperature issues and it's never failed me yet. It does mean that most of my machines don't run at 100% but that's a fair price to pay to run BOINC and still protect the system.

3) Good point. I'll keep that in mind. I did look at SpeedFan many years ago, I recall. I've run every PG project on every machine I've run BOINC on, I'd say, including this laptop very recently. I've never had any issues beyond the odd WU failure, given that the temperature is managed and accepting that WUs won't finish as quickly as they would on more capable, better cooled machines. It's all part of the fun. :-)

I'm thinking about getting around to updating my sadly outdated and disfunctional desktop tower soon, to include some sexy new graphics capabilities. That should be even more fun. ;-)

Cheers.

4) Message boards : Generalized Fermat Prime Search : WUs aborted, restarted, not sure...? (Message 90007)
Posted 1765 days ago by Profile rebel9
... I might try some of the 4-day WUs and see how that works out. Could you also advise which WUs DO trickle?

Thanks.


They all trickle now, but trickles wouldn't have helped you; they merely may have given some insight into what's happening. I suspect in your case it wouldn't have shown anything definitive.

...

To answer your question directly, if you're going to run a GFN task on your computer, you should run the shortest ones -- the n=15 (32768) tasks. Not only are they the shortest tasks, but they're using the x87 transform which consumes less power (and therefore produces less heat.) Those take about 15 minutes on my modern computer and about an hour and a half on my ancient 32-bit AMD Sempron.


OK, thanks Mike. I'll put this one down to experience (Grrrrrrrrrrrr). I appreciate your help. I have now acquired some shorter WUs. They haven't started yet but I'm hopeful they'll be OK.
5) Message boards : Generalized Fermat Prime Search : WUs aborted, restarted, not sure...? (Message 89992)
Posted 1766 days ago by Profile rebel9
If you didn't accidentally omit some lines from the beginning, the constant starting and stopping would appear to be preventing the program from progressing. It never even got to the point where it prints out a time estimate, which should happen within seconds (or at most minutes).


I didn't. It does seem odd but I suspect it's a red herring.


Why is it being stopped and started so often? That's not normal.


I don't know. Perhaps managing CPU load? Suspending when on battery? Unhelpfully, there's no time scale provided in the log.


There are some BOINC settings that could cause that, such as not allowing BOINC to run if the CPU usage is above a certain percentage. There's also third party "thermal" programs that turn BOINC on and off to keep CPU temperatures low.

If you're using either, and *also* have "keep tasks in memory" turned off, the app may not be able to run enough before being turned off, so each time it restarts it will have to start from the beginning.

If this is what's happening, the initial "progress" you saw was actually faked by BOINC because it never got any real progress from the app. No progress was actually occurring. When you restarted, the fake progress indication went back to zero.

Bottom line: check if you have "Leave tasks in memory" turned on. This is a setting on the third tab ("disk and memory usage") of the "computing preferences" dialog in the BOINC manager. If this is turned on, even with constant stopping and starting, the task should still make progress.

Of course, removing the cause of those constant starts and stops would be even better.


I do use a temperature controller. Laptops run far too hot with BOINC otherwise. I killed one in the past. I don't allow BOINC to leave WUs in memory. That causes far too much memory consumption when partially crunched WUs from a stack of projects are left floating around. If you're right, this REALLY sucks. I don't appreciate wasting nearly a month of crunch time. I don't think I'll be bothering with these again. I've just seen the new, extended genefer prefs. Which options would you suggest are best in my circumstances? I might try some of the 4-day WUs and see how that works out. Could you also advise which WUs DO trickle?

Thanks.
6) Message boards : Generalized Fermat Prime Search : WUs aborted, restarted, not sure...? (Message 89980)
Posted 1767 days ago by Profile rebel9
OK, thanks. They're all like this:

genefer 3.2.8 (Windows/CPU/64-bit)
Supported transform implementations: fma3 avx-intel sse4 sse2 default x87
Copyright 2001-2015, Yves Gallot
Copyright 2009, Mark Rodenkirch, David Underbakke
Copyright 2010-2012, Shoichiro Yamada, Ken Brazier
Copyright 2011-2015, Iain Bethune, Michael Goetz, Ronald Schneider
Genefer is free source code, under the MIT license.

Command line: projects/www.primegrid.com/primegrid_genefer_3_2_6_0_3.08_windows_x86_64__cpuGFN.exe -boinc -q 12618^2097152+1

Priority change succeeded.
Using FMA3 transform
Starting initialization...
Initialization complete (11.953 seconds).
Testing 12618^2097152+1...

BOINC client requested that we should suspend.

BOINC client requested that we should resume.

BOINC client requested that we should suspend.

BOINC client requested that we should resume.

BOINC client requested that we should suspend.

BOINC client requested that we should resume.

BOINC client requested that we should suspend.

BOINC client requested that we should resume.

<ad nauseum, then>

BOINC client requested that we should resume.

BOINC client requested that we should suspend.

BOINC client requested that we should resume.

Terminating because BOINC client requested that we should quit.

...which doesn't seem that helpful.
7) Message boards : Generalized Fermat Prime Search : WUs aborted, restarted, not sure...? (Message 89951)
Posted 1768 days ago by Profile rebel9
Hi,

I've been running 4 WUs on my 4-core laptop for about a month. They were all due to finish yesterday. When I checked it late yesterday afternoon, I found that the machine had locked up so I restarted it. On checking BOINC, I found that all 4 WUs appear to have started again from 0. There's no evidence that they were uploaded and that these are new WUs, whch is what I was hoping. The deadline for them is 22/11, which suggests that they're the same ones. Obviously, I don't want to lose a month's credit OR have to sit through the experience again. Is there anything that can be done or is this normal behaviour of some kind? The WUs are:

genefer_2097152_826246
genefer_2097152_826253
genefer_2097152_826250
genefer_2097152_826232

Whilst checking that I've just seen that they are the same WUs as the run times are all ~ 390 hrs. They're all showing ~450hrs est. time remaining. :-(

Thanks.




8) Message boards : Number crunching : New badge proposal - discuss! (Message 64907)
Posted 2700 days ago by Profile rebel9
The other thing is that these are the worst looking badges of any project in the entire BOINC badge-supporting catalogue. While I applaud the current effort, couldn't we take this opportunity to rework them into something slightly bigger, much better looking and representative of the project?

Cheers.
9) Message boards : Project Staging Area : Just checking (Message 57731)
Posted 2925 days ago by Profile rebel9
Hi,

Thanks but getting back to the point. My PSA stats haven't changed since around the end of April. I know from experience that there are several of us using the same user name, so they may not even be all mine, although the machine name is correct, but again may not be unique. None of this explains why my PSA credit has been at 1,194,260.00 since I re-started and not changed. So, my original question again - has there just not been an update or have I made an error in my config., perhaps?

Thanks again.
10) Message boards : Project Staging Area : Just checking (Message 57714)
Posted 2926 days ago by Profile rebel9

Hi Guys,

I started crunching PSA work again a bit over a month ago after a lengthy break but haven't seen any change in my stats. Just checking whether that's correct or if I have something set incorrectly in my config.

Thanks.


Next 10 posts
[Return to PrimeGrid main page]
DNS Powered by DNSEXIT.COM
Copyright © 2005 - 2020 Rytis Slatkevičius (contact) and PrimeGrid community. Server load 3.37, 3.14, 2.72
Generated 25 Sep 2020 | 11:31:41 UTC