Looks like you may need to fix a coupple bad WU's

Message boards : Number crunching : Looks like you may need to fix a coupple bad WU's
Message board moderation

To post messages, you must log in.

AuthorMessage
Grandpa

Send message
Joined: 11 Jun 14
Posts: 11
Credit: 45,376,032
RAC: 0
Message 1275 - Posted: 22 May 2015, 20:44:03 UTC

I noticed a couple of WU's had erred out when checking my work stats they have erred on several other machines besides my own so I am guessing they may be broken.

wu_sf3_DS-10x271_Grp518833of682667 http://numberfields.asu.edu/NumberFields/workunit.php?wuid=9955823

wu_sf3_DS-10x271_Grp491521of682667 http://numberfields.asu.edu/NumberFields/workunit.php?wuid=9902500
ID: 1275 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Eric Driver
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 8 Jul 11
Posts: 1327
Credit: 418,778,706
RAC: 373,208
Message 1276 - Posted: 23 May 2015, 2:15:44 UTC - in response to Message 1275.  

I noticed a couple of WU's had erred out when checking my work stats they have erred on several other machines besides my own so I am guessing they may be broken.

wu_sf3_DS-10x271_Grp518833of682667 http://numberfields.asu.edu/NumberFields/workunit.php?wuid=9955823

wu_sf3_DS-10x271_Grp491521of682667 http://numberfields.asu.edu/NumberFields/workunit.php?wuid=9902500



Thanks for reporting. I have removed those two from the queue.

I am aware of the problem. It happens to about 1 out of 10000 WUs, and is related to factoring the polynomial discriminant. Sometimes the polynomial under test has a very large prime factor in it's discriminant, and the factoring algorithm has problems with it. I end up running those special cases offline using a more robust (but slower) algorithm.

Finding a better way of dealing with this is on my list of things to do, but is a lower priority right now.
ID: 1276 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Grandpa

Send message
Joined: 11 Jun 14
Posts: 11
Credit: 45,376,032
RAC: 0
Message 1277 - Posted: 23 May 2015, 16:31:36 UTC - in response to Message 1276.  
Last modified: 23 May 2015, 16:32:26 UTC

Good to know I will keep an eye out for these when I am running the project, the odds are pretty good that if there are any out there I will get them, since I run so many work units daily when running the project and rarely get errors. Perhaps I can be of some help in keeping the loss of cpu cycles down.

I went back and looked a little further in my history and I think the one listed below may be another one.

wu_sf3_DS-10x271_Grp464214of682667
http://numberfields.asu.edu/NumberFields/workunit.php?wuid=9843193
ID: 1277 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Eric Driver
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 8 Jul 11
Posts: 1327
Credit: 418,778,706
RAC: 373,208
Message 1278 - Posted: 24 May 2015, 1:38:29 UTC - in response to Message 1277.  

Good to know I will keep an eye out for these when I am running the project, the odds are pretty good that if there are any out there I will get them, since I run so many work units daily when running the project and rarely get errors. Perhaps I can be of some help in keeping the loss of cpu cycles down.

I went back and looked a little further in my history and I think the one listed below may be another one.

wu_sf3_DS-10x271_Grp464214of682667
http://numberfields.asu.edu/NumberFields/workunit.php?wuid=9843193


Yes, that was indeed another one. I also queried the database and found a couple more. For some reason they occur in groups. Sometimes I'll see nothing for 100,000 WUs and then there will be a cluster of bad ones.
ID: 1278 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote

Message boards : Number crunching : Looks like you may need to fix a coupple bad WU's


Main page · Your account · Message boards


Copyright © 2024 Arizona State University