| Author |
Message |
|---|
|
|
|
Please can you explain the TRP Sieve tasks I've been sent today
that look as though they will run for longer than 5 months!!!
I've aborted several on 2 different machines that are both set as CPU only
but I will try one out to see just what happens, task below is in progress.
e.g. TRP_2661966_2
Estimated size 38175065 GFLOPS
Est. run time 3782:15:05
Rec'd 25-04-13 - Report D'line 29-04-13...some hope!
Tasks up until today usually take 3.5 hours (ish).
Please advise, best regards
Dave.
____________
|
|
|
 Michael Goetz Volunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13620
ID: 53948
Credit: 266,682,965
RAC: 300,118
                           
|
|
Estimated run times are currently very inaccurate. Try to ignore what BOINC is estimating unless the task is at least 50% completed. Otherwise, just multiply the percentage done by the elapsed time to get a fairly accurate estimate of how long it will take.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Thanks, the task seems to be running okay,
time remaing running down at an amazing rate.
Dave
____________
|
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 2876
ID: 130544
Credit: 1,079,341,085
RAC: 1,299,060
                 
|
|
Yes - time to parade the streets with a big banner "BRING BACK DCF" :D!
Then again that might look a bit strange... |
|
|
|
|
|
I noticed this also today. Boinc estimates 3500 hours to complete! :)
They finish in a couple hours. :)
____________
|
|
|
|
|
I noticed this also today. Boinc estimates 3500 hours to complete! :)
They finish in a couple hours. :)
3500 hours would save a bit in internet costs!!!
____________
Member team AUSTRALIA
My lucky number is 9291*2^1085585+1 |
|
|
|
|
|
Unfortunately, it appears to screw up the work buffer though.
____________
|
|
|
|
|
Unfortunately, it appears to screw up the work buffer though.
Yes this is very true. I'm only running this project on one of my computers so its not a big deal. But since they kick into high priority mode right away (EDIT: and stay that way until they are done) no other task will run. Something did change on the server side. Some are doing it and others are not so I'm thinking a value was changed on the newer work units.
EDIT: Or it's the new version of boinc that just came out, 7.0.65. <--I'm betting on this.
____________
|
|
|
|
|
|
Hate to do this but I'm now dumping any of these
tasks as they're really screwing things up here.
____________
|
|
|
j.sheridanVolunteer tester
Send message
Joined: 21 Mar 11
Posts: 802
ID: 91622
Credit: 1,705,949,285
RAC: 4,532,143
               
|
|
Michael has put a temporary fix in so the tasks are reporting correct times again. |
|
|
|
|
Michael has put a temporary fix in so the tasks are reporting correct times again.
This fix is not working for me. On the machine I am posting this from, the cuda tasks report an initial completion time of 11 hours, but actually finish in 50 minutes. The initial times do not reduce as I report more results.
Could be a BOINC problem.
|
|
|
|
|
Michael has put a temporary fix in so the tasks are reporting correct times again.
This fix is not working for me. On the machine I am posting this from, the cuda tasks report an initial completion time of 11 hours, but actually finish in 50 minutes. The initial times do not reduce as I report more results.
Could be a BOINC problem.
|
|
|
|
|
This fix is not working for me. On the machine I am posting this from, the cuda tasks report an initial completion time of 11 hours, but actually finish in 50 minutes. The initial times do not reduce as I report more results.
Could be a BOINC problem.
It is a BOINC problem because GPU is still not calculated separately from the CPU, so it shows approximately CPU time. Another issue BOINC has is it creates a single Duration Correction Factor per project. Since PrimeGrid has a lot of subprojects under it, it's harder to keep the corrected times spot on for everything.
That's why the backend calculators were made. It still won't show correctly for GPU because again there is only one correction per subproject, currently.
____________
My lucky numbers are 121*2^4553899-1 and 3756801695685*2^666669±1 |
|
|
