Index  | Recent Threads  | Unanswered Threads  | Who's Active  | Guidelines  | Search
 

Quick Go »
No member browsing this thread
Thread Status: Active
Total posts in this thread: 15
Posts: 15   Pages: 2   [ 1 2 | Next Page ]
[ Jump to Last Post ]
Post new Thread
Author
Previous Thread This topic has been viewed 11855 times and has 14 replies Next Thread
BSD
Senior Cruncher
Joined: Apr 27, 2011
Post Count: 224
Status: Offline
Reply to this Post  Reply with Quote 
angry Another stuck workunit

Task Manager shows no CPU activity for any BOINC WU related process, the project application and vina application are still in memory just not showing CPU activity. Nothing in System or Application event viewer logs that would indicate a problem.


Client OS: Windows 7 x86
RAM: 1 GB DDR2
CPU: Intel(R) Pentium(R) M processor 1300MHz
BOINC Mgr Processor Use at Most: 50%


WU: GFAM_x1kmvHumanDHFRdry_0001468_0741_0

Progress: 75.416%
CPU Time: 09:30:00
Elapsed Time: 35:20:46 shock d oh crying angry
App: 6.08


Will suspend WU task, reboot computer, and resume WU task to get it going again. More wasted time and electricty.

What a waste money eyes

Edit: Added CPU % use and spelling.
----------------------------------------
[Edit 2 times, last edit by BSD at Dec 12, 2011 3:03:43 PM]
[Dec 12, 2011 2:45:50 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Another stuck workunit

Hello BSD,
This sort of problem does not happen on my computer. You should check your profile settings and see if there are any that might tell BOINC to stop running your work unit.

Lawrence
[Dec 12, 2011 6:18:49 PM]   Link   Report threatening or abusive post: please login first  Go to top 
BSD
Senior Cruncher
Joined: Apr 27, 2011
Post Count: 224
Status: Offline
Reply to this Post  Reply with Quote 
Re: Another stuck workunit

It's the "Work" profile with 09:00 to 10:00 network time available and 50% limit CPU use set in the local override. Nothing else I can see that would aversely affect crunching and cause a WU to stop processing.

Some mysterious set of circumstances: sun spots, lunar eclipse, fluctuation in the Earth's magnetic field? silly Sorry, just trying to be humorous and not worry about it too much.

I've had similar stuck WU on another device, different device profile, OS, and CPU. That was about a month or so ago.
[Dec 12, 2011 8:30:03 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Another stuck workunit

As with other sciences such as DSFL/GFAM the suspicion is on the BOINC Throttle when observing stuck units [demonstrated by a member who had the problem with DSFL and had it go away by running 100%]. Put that control back to 100% and when on Windows use ThreadMasterGUI to set a % control on a "per science" basis. I use that as different sciences generate less or more heat. E.g. during the summer got C4CW on 80%, but HFCC to 90% on my laptop. Not all calculations load a the different functions of CPU equally.

--//--

edit: Also if temperature is a concern use TThrottle for automated crunching reduction if it gets to hot. Set the ceiling to for instance 70C and whenever that is exceeded, BOINC is slowed down, not paused, to get the temps down.
----------------------------------------
[Edit 1 times, last edit by Former Member at Dec 12, 2011 10:38:07 PM]
[Dec 12, 2011 10:36:27 PM]   Link   Report threatening or abusive post: please login first  Go to top 
BSD
Senior Cruncher
Joined: Apr 27, 2011
Post Count: 224
Status: Offline
Reply to this Post  Reply with Quote 
Re: Another stuck workunit

As with other sciences such as DSFL/GFAM the suspicion is on the BOINC Throttle when observing stuck units [demonstrated by a member who had the problem with DSFL and had it go away by running 100%].

I have 6 devices with BOINC set to 50% CPU, been recently running DSFL/GFAM for about a week now after ~2 months crunching other projects.

This stuck WU was running on an older IBM R40 laptop (Intel CPU), that device had another stuck WU several months ago but don't remember the exact details.

I have another older Dell Vostro 1000 laptop (AMD dual CPU) that recently had two GFAM WUs get stuck at the same time.

The other 4 devices are desktops (AMD and Intel multi-core CPUs), to the best of recollection, none of them have had stuck WUs. So maybe, at least in my case, the "suspicion" is a "mobile" CPU as opposed to a "desktop" CPU? I have 3 more devices running DSFL/GFAM, but those have been running at 100% CPU with no stuck WUs.

p.s. Just remembered I have new Dell Latitute laptop (Intel 4 core with HT on, 8 threads) and as of yet haven't seen a stuck WU, but it is running at 100% CPU. Just lucky I guess.

Put that control back to 100% and when on Windows use ThreadMasterGUI to set a % control on a "per science" basis.

Up, then back down? I must admit, I don't understand "where" the change occurs? By putting the BOINC Manager up to 100%, is the "suspicion" a programing bug in BOINC as opposed to the science application?

Thanks SekeRob for the information. peace
[Dec 13, 2011 12:37:55 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
GFAM: Another stuck workunit

Thing is, BOINC's throttle is crude, run 1 second, pause 1 second at 50% [to cool down, it's only design purpose]. TMG will just reduce the amount of cycles that the science apps get. Been using this tool for probably 5 years if not longer. Never seen a stuck result... on my laptop [Intel Centrino]. :D

The TMG tool has a topic in the Start Here FAQ index. TThrottle [specially written for BOINC on Windows], does it similar, just using a temperature ceiling you can set and slowing the cycles that get through to BOINC. The tool works for both CPU and GPU.

--//--
----------------------------------------
[Edit 1 times, last edit by Former Member at Dec 13, 2011 7:40:30 AM]
[Dec 13, 2011 7:38:55 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: GFAM: Another stuck workunit

I suspect that either you have bad memory or not enouth to run your system and the task. No problems with my dell laptops. But the all have at least 8 gig mem and are xps systems.
[Dec 13, 2011 11:12:08 AM]   Link   Report threatening or abusive post: please login first  Go to top 
BSD
Senior Cruncher
Joined: Apr 27, 2011
Post Count: 224
Status: Offline
Reply to this Post  Reply with Quote 
Re: GFAM: Another stuck workunit

I suspect that either you have bad memory or not enouth to run your system and the task. No problems with my dell laptops. But the all have at least 8 gig mem and are xps systems.

Been running this IBM R40 laptop in it's current configuration and crunching all types of WUs, even CEP2, for over a year now without problems except for recent stuck WUs. Maybe it's time to put it out to pasture.

Do you run all your laptops at 100% CPU? Do you use the TThrottle app?
[Dec 13, 2011 11:55:51 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: GFAM: Another stuck workunit

i run them all at 100% and never use tthrottle on my smallest laptop xps dual core i run it at 50% processor but 100% cpu time due to heat.
[Dec 13, 2011 12:14:57 PM]   Link   Report threatening or abusive post: please login first  Go to top 
mikey
Veteran Cruncher
Joined: May 10, 2009
Post Count: 821
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: GFAM: Another stuck workunit

i run them all at 100% and never use tthrottle on my smallest laptop xps dual core i run it at 50% processor but 100% cpu time due to heat.


Did you stop Boinc and then restart it? If so did the unit start back up again? For some obscure reason Boinc just stops processing some times and it happens so infrequently it is hard to track down. A restart of Boinc, or sometimes a restart of the pc, usually fixes the problem. Yes it could be memory, it could even be mismatched memory but it is a puzzler, but is often okay after a restart of Boinc.
----------------------------------------


[Dec 13, 2011 1:01:00 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Posts: 15   Pages: 2   [ 1 2 | Next Page ]
[ Jump to Last Post ]
Post new Thread