So how long is your C116 Task?

Message boards : Number crunching : So how long is your C116 Task?
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
Werinbert

Send message
Joined: 9 May 13
Posts: 10
Credit: 100,312
RAC: 0
United States
Message 683 - Posted: 15 Aug 2014, 15:23:03 UTC

Just got one back that took longer than expected.

C116 73,474.32sec on an i5 2410M@2.3GHz/3 cores

This is pushing the limit of how long I want to tie-up my computer (I would prefer less than 10 hours). The Yafu task pushed aside three PrimeGrid tasks which will make it back on time... if only barely.
ID: 683 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile yoyo_rkn
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist

Send message
Joined: 22 Aug 11
Posts: 725
Credit: 16,445,605
RAC: 5
Germany
Message 684 - Posted: 16 Aug 2014, 5:45:06 UTC - in response to Message 683.  

Thanks for posting the runtimes.
ID: 684 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Bruce Kennedy

Send message
Joined: 8 Sep 11
Posts: 4
Credit: 1,048,224
RAC: 0
United States
Message 722 - Posted: 7 Oct 2014, 3:38:15 UTC

I have a C116 that has been running for 24 hours now using all 8 cores and is at 99.999% for the last 4 hours or so. I appears the CPU time is stuck also at 03:25:31. Will it ever finish on my i7-4700HQ? If it's not done in the next 12 hours or so I'll cancel it.
ID: 722 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile yoyo_rkn
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist

Send message
Joined: 22 Aug 11
Posts: 725
Credit: 16,445,605
RAC: 5
Germany
Message 723 - Posted: 7 Oct 2014, 5:57:39 UTC - in response to Message 722.  

You will get credits also after the deadline, so don't worry.

Can you post the content of the slot folder and check when it was changed the last time?

yoyo
ID: 723 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Bruce Kennedy

Send message
Joined: 8 Sep 11
Posts: 4
Credit: 1,048,224
RAC: 0
United States
Message 724 - Posted: 7 Oct 2014, 9:17:03 UTC - in response to Message 723.  

The C116 WU is now at 100% and waiting to run. Here are the contents of the text files in the slot directory.

10/05/14 22:50:24 v1.34.5 @ BRUCE,
10/05/14 22:50:24 v1.34.5 @ BRUCE, ****************************
10/05/14 22:50:24 v1.34.5 @ BRUCE, Starting factorization of 38248148288262939877592328434712626897417157843496786404118652121030605576532119227866057116701426977912061693502551
10/05/14 22:50:24 v1.34.5 @ BRUCE, using pretesting plan: normal
10/05/14 22:50:24 v1.34.5 @ BRUCE, no tune info: using qs/gnfs crossover of 95 digits
10/05/14 22:50:24 v1.34.5 @ BRUCE, ****************************
10/05/14 22:50:24 v1.34.5 @ BRUCE, rho: x^2 + 3, starting 1000 iterations on C116
10/05/14 22:50:25 v1.34.5 @ BRUCE, rho: x^2 + 2, starting 1000 iterations on C116
10/05/14 22:50:25 v1.34.5 @ BRUCE, rho: x^2 + 1, starting 1000 iterations on C116
10/05/14 22:50:25 v1.34.5 @ BRUCE, pm1: starting B1 = 150K, B2 = gmp-ecm default on C116
10/05/14 22:50:25 v1.34.5 @ BRUCE, current ECM pretesting depth: 0.00
10/05/14 22:50:25 v1.34.5 @ BRUCE, scheduled 30 curves at B1=2000 toward target pretesting depth of 35.69
10/05/14 22:50:26 v1.34.5 @ BRUCE, Finished 30 curves using Lenstra ECM method on C116 input, B1=2K, B2=gmp-ecm default
10/05/14 22:50:26 v1.34.5 @ BRUCE, current ECM pretesting depth: 15.18
10/05/14 22:50:26 v1.34.5 @ BRUCE, scheduled 74 curves at B1=11000 toward target pretesting depth of 35.69
10/05/14 22:50:41 v1.34.5 @ BRUCE, Finished 74 curves using Lenstra ECM method on C116 input, B1=11K, B2=gmp-ecm default
10/05/14 22:50:41 v1.34.5 @ BRUCE, current ECM pretesting depth: 20.24
10/05/14 22:50:41 v1.34.5 @ BRUCE, scheduled 214 curves at B1=50000 toward target pretesting depth of 35.69
10/05/14 22:52:44 v1.34.5 @ BRUCE, Finished 214 curves using Lenstra ECM method on C116 input, B1=50K, B2=gmp-ecm default
10/05/14 22:52:44 v1.34.5 @ BRUCE, pm1: starting B1 = 3750K, B2 = gmp-ecm default on C116
10/05/14 22:52:47 v1.34.5 @ BRUCE, current ECM pretesting depth: 25.33
10/05/14 22:52:47 v1.34.5 @ BRUCE, scheduled 430 curves at B1=250000 toward target pretesting depth of 35.69
10/05/14 23:07:07 v1.34.5 @ BRUCE, Finished 430 curves using Lenstra ECM method on C116 input, B1=250K, B2=gmp-ecm default
10/05/14 23:07:07 v1.34.5 @ BRUCE, pm1: starting B1 = 15M, B2 = gmp-ecm default on C116
10/05/14 23:07:18 v1.34.5 @ BRUCE, current ECM pretesting depth: 30.45
10/05/14 23:07:18 v1.34.5 @ BRUCE, scheduled 904 curves at B1=1000000 toward target pretesting depth of 35.69
10/06/14 01:26:07 v1.34.5 @ BRUCE, Finished 904 curves using Lenstra ECM method on C116 input, B1=1M, B2=gmp-ecm default
10/06/14 01:26:07 v1.34.5 @ BRUCE, current ECM pretesting depth: 35.56
10/06/14 01:26:07 v1.34.5 @ BRUCE, scheduled 66 curves at B1=3000000 toward target pretesting depth of 35.69
10/06/14 01:54:48 v1.34.5 @ BRUCE, Finished 66 curves using Lenstra ECM method on C116 input, B1=3M, B2=gmp-ecm default
10/06/14 01:54:48 v1.34.5 @ BRUCE, final ECM pretested depth: 35.70
10/06/14 01:54:48 v1.34.5 @ BRUCE, scheduler: switching to sieve method
10/06/14 01:54:48 v1.34.5 @ BRUCE, nfs: commencing nfs on c116: 38248148288262939877592328434712626897417157843496786404118652121030605576532119227866057116701426977912061693502551
10/06/14 01:54:48 v1.34.5 @ BRUCE, nfs: commencing poly selection with 1 threads
10/06/14 01:54:48 v1.34.5 @ BRUCE, nfs: setting deadline of 7650 seconds
10/06/14 03:59:15 v1.34.5 @ BRUCE, nfs: completed 44 ranges of size 250 in 7466.9067 seconds
10/06/14 03:59:15 v1.34.5 @ BRUCE, nfs: best poly = # norm 4.575890e-011 alpha -5.879554 e 4.598e-010 rroots 3
10/06/14 03:59:15 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 05:11:38 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 06:25:39 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 07:34:46 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 08:49:54 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 10:00:20 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 11:16:24 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 12:30:38 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 13:43:13 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 15:04:15 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 16:20:10 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 17:31:18 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 18:45:41 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 19:58:42 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 21:10:20 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 22:26:56 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/06/14 23:45:59 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/07/14 00:57:21 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads
10/07/14 02:17:26 v1.34.5 @ BRUCE, nfs: commencing lattice sieving with 1 threads

LatSieveTime: 4329
LatSieveTime: 4425
LatSieveTime: 4129
LatSieveTime: 4493
LatSieveTime: 4205
LatSieveTime: 4546
LatSieveTime: 4436
LatSieveTime: 4339
LatSieveTime: 4842
LatSieveTime: 4536
LatSieveTime: 4251
LatSieveTime: 4440
LatSieveTime: 4363
LatSieveTime: 4281
LatSieveTime: 4575
LatSieveTime: 4724
LatSieveTime: 4266
LatSieveTime: 4775
LatSieveTime: 4860

0 12348.187500
ID: 724 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile yoyo_rkn
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist

Send message
Joined: 22 Aug 11
Posts: 725
Credit: 16,445,605
RAC: 5
Germany
Message 725 - Posted: 7 Oct 2014, 12:29:10 UTC - in response to Message 724.  

It runs only with 1 thread.
There was a short problem on the server which issued workunits only with 1 thread. So it will take longer to finish this workunit.
ID: 725 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Bruce Kennedy

Send message
Joined: 8 Sep 11
Posts: 4
Credit: 1,048,224
RAC: 0
United States
Message 726 - Posted: 8 Oct 2014, 12:51:47 UTC - in response to Message 725.  

It may be using 1 thread but it has all 8 tied up and prevents other work from being done. I'm going to cancel it now, it's be running for 36 hours.
ID: 726 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Beyond

Send message
Joined: 4 Oct 14
Posts: 36
Credit: 131,624,774
RAC: 57,558
United States
Message 730 - Posted: 19 Nov 2014, 16:24:12 UTC
Last modified: 19 Nov 2014, 16:29:06 UTC

Hi yoyo,

I have a WU that's been running for 132 hours and has been at 100% for the last 2 days:

http://yafu.myfirewall.org/yafu/workunit.php?wuid=227630

It's on a slow 2 core Ivy Bridge box, but the longest previous WU on this machine was 21.3 hours. Should I leave it run or is it dead? It's still using CPU, but looks like only 1 core.

It seemed to run fairly normally at first, but then as the completion percentage got higher it went slower and slower. just before hitting 100% it was progressing at about 0.001% per 2 hours.

Edit: Here's the output from a log file if that helps.:

11/13/14 22:13:02 v1.34.5 @ SUBA,
11/13/14 22:13:02 v1.34.5 @ SUBA, ****************************
11/13/14 22:13:02 v1.34.5 @ SUBA, Starting factorization of 47987412564198200109186934880570980526987301425001440780338916116510420217531104751755815639270422743607663313229559
11/13/14 22:13:02 v1.34.5 @ SUBA, using pretesting plan: normal
11/13/14 22:13:02 v1.34.5 @ SUBA, no tune info: using qs/gnfs crossover of 95 digits
11/13/14 22:13:02 v1.34.5 @ SUBA, ****************************
11/13/14 22:13:02 v1.34.5 @ SUBA, rho: x^2 + 3, starting 1000 iterations on C116
11/13/14 22:13:02 v1.34.5 @ SUBA, rho: x^2 + 2, starting 1000 iterations on C116
11/13/14 22:13:02 v1.34.5 @ SUBA, rho: x^2 + 1, starting 1000 iterations on C116
11/13/14 22:13:02 v1.34.5 @ SUBA, pm1: starting B1 = 150K, B2 = gmp-ecm default on C116
11/13/14 22:13:02 v1.34.5 @ SUBA, current ECM pretesting depth: 0.00
11/13/14 22:13:02 v1.34.5 @ SUBA, scheduled 30 curves at B1=2000 toward target pretesting depth of 35.69
11/13/14 22:13:02 v1.34.5 @ SUBA, Finished 30 curves using Lenstra ECM method on C116 input, B1=2K, B2=gmp-ecm default
11/13/14 22:13:02 v1.34.5 @ SUBA, current ECM pretesting depth: 15.18
11/13/14 22:13:02 v1.34.5 @ SUBA, scheduled 74 curves at B1=11000 toward target pretesting depth of 35.69
11/13/14 22:13:10 v1.34.5 @ SUBA, Finished 74 curves using Lenstra ECM method on C116 input, B1=11K, B2=gmp-ecm default
11/13/14 22:13:10 v1.34.5 @ SUBA, current ECM pretesting depth: 20.24
11/13/14 22:13:10 v1.34.5 @ SUBA, scheduled 214 curves at B1=50000 toward target pretesting depth of 35.69
11/13/14 22:14:23 v1.34.5 @ SUBA, Finished 214 curves using Lenstra ECM method on C116 input, B1=50K, B2=gmp-ecm default
11/13/14 22:14:23 v1.34.5 @ SUBA, pm1: starting B1 = 3750K, B2 = gmp-ecm default on C116
11/13/14 22:14:26 v1.34.5 @ SUBA, current ECM pretesting depth: 25.33
11/13/14 22:14:26 v1.34.5 @ SUBA, scheduled 430 curves at B1=250000 toward target pretesting depth of 35.69
11/13/14 22:22:14 v1.34.5 @ SUBA, Finished 430 curves using Lenstra ECM method on C116 input, B1=250K, B2=gmp-ecm default
11/13/14 22:22:14 v1.34.5 @ SUBA, pm1: starting B1 = 15M, B2 = gmp-ecm default on C116
11/13/14 22:22:25 v1.34.5 @ SUBA, current ECM pretesting depth: 30.45
11/13/14 22:22:25 v1.34.5 @ SUBA, scheduled 904 curves at B1=1000000 toward target pretesting depth of 35.69
11/13/14 23:24:54 v1.34.5 @ SUBA, Finished 904 curves using Lenstra ECM method on C116 input, B1=1M, B2=gmp-ecm default
11/13/14 23:24:54 v1.34.5 @ SUBA, current ECM pretesting depth: 35.56
11/13/14 23:24:54 v1.34.5 @ SUBA, scheduled 66 curves at B1=3000000 toward target pretesting depth of 35.69
11/13/14 23:37:18 v1.34.5 @ SUBA, Finished 66 curves using Lenstra ECM method on C116 input, B1=3M, B2=gmp-ecm default
11/13/14 23:37:18 v1.34.5 @ SUBA, final ECM pretested depth: 35.70
11/13/14 23:37:18 v1.34.5 @ SUBA, scheduler: switching to sieve method
11/13/14 23:37:18 v1.34.5 @ SUBA, nfs: commencing nfs on c116: 47987412564198200109186934880570980526987301425001440780338916116510420217531104751755815639270422743607663313229559
11/13/14 23:37:18 v1.34.5 @ SUBA, nfs: commencing poly selection with 2 threads
11/13/14 23:37:18 v1.34.5 @ SUBA, nfs: setting deadline of 4050 seconds
11/14/14 00:44:48 v1.34.5 @ SUBA, nfs: completed 54 ranges of size 250 in 4050.3227 seconds
11/14/14 00:44:48 v1.34.5 @ SUBA, nfs: best poly = # norm 5.194789e-011 alpha -5.928782 e 5.000e-010 rroots 3
11/14/14 00:44:48 v1.34.5 @ SUBA, nfs: commencing lattice sieving with 2 threads
11/14/14 01:31:26 v1.34.5 @ SUBA, nfs: commencing lattice sieving with 2 threads
11/14/14 02:16:38 v1.34.5 @ SUBA, nfs: commencing lattice sieving with 2 threads
11/14/14 02:59:16 v1.34.5 @ SUBA, nfs: commencing lattice sieving with 2 threads
ID: 730 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile yoyo_rkn
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist

Send message
Joined: 22 Aug 11
Posts: 725
Credit: 16,445,605
RAC: 5
Germany
Message 731 - Posted: 19 Nov 2014, 18:17:10 UTC - in response to Message 730.  

The log looks fine. But it stops on 14. November. So from the log it seems the wu has stopped. Don't trust the progress percentage. This is not from the application. This is some assumption of Boinc and boinctasks.

Please check the slot directory which files were recently changed.

There should be a nfs.dat file which goes up to 1 GB.

yoyo
ID: 731 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Beyond

Send message
Joined: 4 Oct 14
Posts: 36
Credit: 131,624,774
RAC: 57,558
United States
Message 732 - Posted: 19 Nov 2014, 19:08:24 UTC - in response to Message 731.  
Last modified: 19 Nov 2014, 19:09:05 UTC

The nfs.dat in the slot directory is only 116,000K and hasn't been updated since 11/14. I suppose it's dead. I was assuming that it was still running because of the BOINC progress percentage and since the Yafu task still shows as using 50% of the CPU in task manager. I just paused the WU. Do you want the nfs.dat sent to you?
ID: 732 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile yoyo_rkn
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist

Send message
Joined: 22 Aug 11
Posts: 725
Credit: 16,445,605
RAC: 5
Germany
Message 733 - Posted: 19 Nov 2014, 22:17:26 UTC - in response to Message 732.  

No need to send it. Just abort the wu.

yoyo
ID: 733 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Beyond

Send message
Joined: 4 Oct 14
Posts: 36
Credit: 131,624,774
RAC: 57,558
United States
Message 734 - Posted: 20 Nov 2014, 3:46:56 UTC

Aborted. Thanks yoyo!
ID: 734 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Beyond

Send message
Joined: 4 Oct 14
Posts: 36
Credit: 131,624,774
RAC: 57,558
United States
Message 735 - Posted: 21 Nov 2014, 20:41:36 UTC

Now I have one on a different machine that's been sitting at 100% for hours. It's till using 100% CPU on 2 cores. The only files updating since yesterday are .last_spq0 and .last_spq1 which update every minute. Everything else in the slot directory is about 20 hours old. Kill it or keep running it?
ID: 735 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile yoyo_rkn
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist

Send message
Joined: 22 Aug 11
Posts: 725
Credit: 16,445,605
RAC: 5
Germany
Message 736 - Posted: 21 Nov 2014, 22:11:05 UTC - in response to Message 735.  

Which process is consuming CPU?
I would say let it run.
ID: 736 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Beyond

Send message
Joined: 4 Oct 14
Posts: 36
Credit: 131,624,774
RAC: 57,558
United States
Message 737 - Posted: 22 Nov 2014, 14:16:35 UTC - in response to Message 736.  
Last modified: 22 Nov 2014, 14:20:17 UTC

I paused it to let 2 other yafu WUs run that were approaching deadline. Restarted it a few hours ago and just now it aborted with a computation error (196):

11-22-14 08:04 Aborting task yafu_C101_F1416456605_312_0: exceeded disk limit

From Stderr:

<core_client_version>7.4.27</core_client_version>
<![CDATA[
<message>
Maximum disk usage exceeded
</message>
<stderr_txt>
al yield: 6034233, q=15083179 (0.00956 sec/rel)

http://yafu.myfirewall.org/yafu/result.php?resultid=237720
ID: 737 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile yoyo_rkn
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist

Send message
Joined: 22 Aug 11
Posts: 725
Credit: 16,445,605
RAC: 5
Germany
Message 738 - Posted: 22 Nov 2014, 16:53:55 UTC - in response to Message 737.  

Any Idea what file did the disk limit exeed?
ID: 738 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Beyond

Send message
Joined: 4 Oct 14
Posts: 36
Credit: 131,624,774
RAC: 57,558
United States
Message 739 - Posted: 22 Nov 2014, 20:50:00 UTC - in response to Message 738.  

No, when I saw it had aborted I looked in the slot dir and it was empty. The disk has 34GB free. I looked at the slot dir about an hour earlier and still the only files updating were the 2 mentioned above. On all my other Yafu WUs there are a number of files that update perhaps every 10 minutes. Also. what's the maximum ram the yafu app uses?
ID: 739 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile yoyo_rkn
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist

Send message
Joined: 22 Aug 11
Posts: 725
Credit: 16,445,605
RAC: 5
Germany
Message 740 - Posted: 22 Nov 2014, 22:26:21 UTC - in response to Message 739.  

The settings for the wus are:
Max Memory Usage 500 000 000
Max Disk Usage 4000 000 000

This was never reached for C116. I only saw the nfs.dat in the slot folder grow up tu 1 GB.
ID: 740 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Beyond

Send message
Joined: 4 Oct 14
Posts: 36
Credit: 131,624,774
RAC: 57,558
United States
Message 744 - Posted: 26 Nov 2014, 0:12:41 UTC

Creditnew is bizarre. Three identical machines, scoring for same length WUs as much as 8x difference, actually sometimes more. Have no idea why.
ID: 744 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile yoyo_rkn
Volunteer moderator
Project administrator
Project developer
Project tester
Volunteer developer
Volunteer tester
Project scientist

Send message
Joined: 22 Aug 11
Posts: 725
Credit: 16,445,605
RAC: 5
Germany
Message 745 - Posted: 26 Nov 2014, 8:47:13 UTC - in response to Message 744.  

Same length of a wu, e.g. C116, doesn't mean that the runtime is the same. For one WU it might happen, that factors are found direct aft start. For another wu it might be that the checking for known prime factores and trial factoring with ecm doesn't find anything and a complete nfs must be done.
So runtime can differ from some seconds to some hours, for the same number length.

yoyo
ID: 745 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
1 · 2 · Next

Message boards : Number crunching : So how long is your C116 Task?




Datenschutz / Privacy Copyright © 2011-2024 Rechenkraft.net e.V. & yoyo