Advanced search

Message boards : Number Crunching : Problem with WU's?

Author Message
Larry
Send message
Joined: 18 Mar 15
Posts: 2
Combined Credit: 12,359
DNA@Home: 12,359
SubsetSum@Home: 0
Wildlife@Home: 0
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 0

  
Message 5802 - Posted: 22 Aug 2015, 20:54:45 UTC

Is there a problem with the current batch of WU's of which gibbs_snail_hg19_1000fa_3motifs_6width_norepeats_2_201_252770_150000_0 is/was a member of?

This WU and a previous WU ending in _252715_190000_0 uncharacteristically thrashed my HDD for 4+ hours with essentially no (<0.5%) progress prior to my aborting both tasks.

Please advise as I have configured "no new tasks" for this project in the meantime.

Michal Kindura
Send message
Joined: 28 Apr 13
Posts: 123
Combined Credit: 81,669,569
DNA@Home: 253,257
SubsetSum@Home: 1,127,118
Wildlife@Home: 80,289,195
Wildlife@Home Watched: 38,247,903s
Wildlife@Home Events: 12,091
Climate Tweets: 4
Images Observed: 3,195

            
Message 5803 - Posted: 23 Aug 2015, 2:44:05 UTC - in response to Message 5802.

Is there a problem with the current batch of WU's of which gibbs_snail_hg19_1000fa_3motifs_6width_norepeats_2_201_252770_150000_0 is/was a member of?

This WU and a previous WU ending in _252715_190000_0 uncharacteristically thrashed my HDD for 4+ hours with essentially no (<0.5%) progress prior to my aborting both tasks.

Please advise as I have configured "no new tasks" for this project in the meantime.

You have only 512MB RAM, I see. Is it right? I think the problem is in the memory. One DNA task may require up to 800MB of RAM.

It's just my experience. The problem may be elsewhere.

Larry
Send message
Joined: 18 Mar 15
Posts: 2
Combined Credit: 12,359
DNA@Home: 12,359
SubsetSum@Home: 0
Wildlife@Home: 0
Wildlife@Home Watched: 0s
Wildlife@Home Events: 0
Climate Tweets: 0
Images Observed: 0

  
Message 5805 - Posted: 23 Aug 2015, 17:38:04 UTC - in response to Message 5803.

This is true, and would account for all the HDD thrashing. And, typically I receive WU's of much shorter estimated duration (smaller size and lower requirements) on this machine.

It's just the combination of "uncharacteristic" drive thrashing and lack of "expected" progress based on run-time vs estimated-remaining that raised my concern of these 2 [sequentially received] jobs.

I'll re-open the gate and see what I get this time. Thanks for your input Michal.

Michal Kindura
Send message
Joined: 28 Apr 13
Posts: 123
Combined Credit: 81,669,569
DNA@Home: 253,257
SubsetSum@Home: 1,127,118
Wildlife@Home: 80,289,195
Wildlife@Home Watched: 38,247,903s
Wildlife@Home Events: 12,091
Climate Tweets: 4
Images Observed: 3,195

            
Message 5809 - Posted: 24 Aug 2015, 18:07:30 UTC - in response to Message 5805.
Last modified: 24 Aug 2015, 18:08:07 UTC

The hard drive could fail because - swap.

If Windows XP don't have enough RAM, furiously begin to write to the pagefile.sys (swap, virtual memory).

I don't know the details about DNA workunits. Again, only based on my experiences... There are 3 types of WU:

Extreme short (10*n MB in RAM), computing time ~ 5 minutes.
Medium units (~200MB in RAM), computing time ~ 40 minutes.
Extreme long (800-1200MB in RAM), computing time ~ 3 hours.

You can't choose, or only I can't found these settings.

Travis Desell
Volunteer moderator
Project administrator
Project developer
Project scientist
Send message
Joined: 16 Jan 12
Posts: 1813
Combined Credit: 23,514,257
DNA@Home: 293,563
SubsetSum@Home: 349,212
Wildlife@Home: 22,871,482
Wildlife@Home Watched: 212,926s
Wildlife@Home Events: 51
Climate Tweets: 21
Images Observed: 774

              
Message 5825 - Posted: 29 Aug 2015, 7:11:43 UTC - in response to Message 5809.

The RAM issue is most likely the case. I can see about upping the memory limit on the workunits so this doesn't happen.


Post to thread

Message boards : Number Crunching : Problem with WU's?