BOINC 6.6.38 - No WU progress on one Project.

Message boards : Questions and problems : BOINC 6.6.38 - No WU progress on one Project.
Message board moderation

To post messages, you must log in.

AuthorMessage
Profile jay_e

Send message
Joined: 8 Mar 07
Posts: 115
United States
Message 28173 - Posted: 20 Oct 2009, 21:40:02 UTC

Greetings!
I'm a happy cruncher and recently added the Malaria control project on a PC with Windows XP Pro.

The "time to complete" data kept incrementing and the WU self-aborted with "Maximum elapsed time exceeded"

I tried a lot of things - even an uninstall and re-install of BOINC. Still got error.

Tried without AV running - still got error.

Tried with no other WU from any other project - still got error,

Tried on another PC - Windows XP home; different AV - still got error.

WU from other projects work fine.
WU sent to wingman finished in ~3 hours - OK.


Took Off gloves. Finished all WU. Detached all projects. Uninstalled Boinc. Deleted Persistent data in 'application data' directory. Used regedit to scan and delete remnants as needed.

Ran Performance tests (FPU) - Passed.

Ram Memory tests - Passed.

Get fresh copy of BOINC 6.6.38
Attach back to MC as existing user.
Get WU and observe progress.

Jubilation as WU started with a "To completion code"
of 03:06:16 and is started DECREMENTING.

And then.. frustration. Looks like problem is still there.

Elapsed time is 01:10:00
To completion time is 03:12:04
Progress is 20.89%

So far it looks like the problem is back.
I'll wait the 25 hours on 100% CPU on theis WU and see what happens.


--
So. The question is: has anyone else seen this behavior?
Ideas? Love letters? Advice?

Thanks,
Jay

ps
See
http://www.malariacontrol.net/forum_thread.php?id=918

for details...

Thanks again,
Jay

ID: 28173 · Report as offensive
Profile Jord
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 15490
Netherlands
Message 28175 - Posted: 20 Oct 2009, 21:44:06 UTC - in response to Message 28173.  

"Maximum elapsed time exceeded" is a science application error. You will need to report this on the forums of the affected project. Their <rsc_fpops_bound> value on their tasks is too low then.
ID: 28175 · Report as offensive
Profile jay_e

Send message
Joined: 8 Mar 07
Posts: 115
United States
Message 28181 - Posted: 20 Oct 2009, 22:37:52 UTC - in response to Message 28175.  

Hi Jord,

Here is the problem.
No one else in MC has this problem.
Similar PCs finsh the WU in three hours.

Yes, I see that this is an MC problem - but I wanted to see if any other project had the problem - and solved it.

I'll see if they have a linux app and try that too....


Optimistic and hopeful,
Jay
ID: 28181 · Report as offensive
Profile Jord
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 15490
Netherlands
Message 28182 - Posted: 20 Oct 2009, 22:45:33 UTC - in response to Message 28181.  

If no one else has the problem, then it's usually hardware related. Something in your computer stopping, probably the harddrive (is it set to stop spinning after x-amount of time to be greener?), the CPU or memory.


ID: 28182 · Report as offensive
Profile jay_e

Send message
Joined: 8 Mar 07
Posts: 115
United States
Message 28184 - Posted: 21 Oct 2009, 1:30:18 UTC - in response to Message 28182.  

If no one else has the problem, then it's usually hardware related. Something in your computer stopping, probably the harddrive (is it set to stop spinning after x-amount of time to be greener?), the CPU or memory.



I ran CPU, FPU and memory tests. All were OK

Power setting? Ah-Hah! That might be it.
The power option was set to turn off hard disk after one hour.
I thought that meant the hard disk had to be idle for an hour.
MC uses the disk quite a bit - and I thought the disk would never shut off.

Also after detaching all projects; uninstalling BOINC; scrubbing directories; cleaning with regedit; re-installing BOINC; and re-loading the project, One wu has completed in 3.75 hours.

((((Hurray!!))))


I'll leave the power setting as is overnight; keep my fingers off;
and see if the remaining 5 wu complete.
If there is a problem, I'll set the disk to always on.

If that's not it, I'll try changing the cpu throttle to 50% and then back to 100% - to see if that re-creates the problem.

Thanks,
Jay
ID: 28184 · Report as offensive

Message boards : Questions and problems : BOINC 6.6.38 - No WU progress on one Project.

Copyright © 2024 University of California.
Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.