Message boards : Sixtrack Application : SIXTRACKTEST
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 3 · 4 · 5 · 6 · 7 · 8 · Next

AuthorMessage
Alessio Mereghetti
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 29 Feb 16
Posts: 141
Credit: 1,744,986
RAC: 5,096
Message 37633 - Posted: 18 Dec 2018, 9:08:25 UTC - in response to Message 37632.  
Last modified: 18 Dec 2018, 9:10:58 UTC

Hello Crystal Pellet,
many thanks for the feedback. Yes, I have experienced similar problems on my Linux desktop pc.
As you have figured out on your own, the big beast is mem_alloc.log, a debug file we use to monitor dynamic array allocation. We forgot to put it under DEBUG pre-processing flag, and we abused of dynamic array allocation in a specific routine for beam-beam effects. We have fixed the two issues already yesterday, but we wait for other stuff to come up before updating the exes. In the meanwhile, I am submitting jobs with v5.01.01 without beam-beam effects.

We are also testing version pinning, so that's why you may see arriving sixtracktest tasks with v4.6.30.

I am a bit more interested by your successful tasks, e.g.
https://lhcathome.cern.ch/lhcathome/result.php?resultid=212498881
https://lhcathome.cern.ch/lhcathome/result.php?resultid=212498882
https://lhcathome.cern.ch/lhcathome/result.php?resultid=212498883

They ran for quite a while (i.e. >1h), and I don't understand how it is possible that they did not hit the disk barrier...
ID: 37633 · Report as offensive     Reply Quote
Crystal Pellet
Volunteer moderator
Volunteer tester

Send message
Joined: 14 Jan 10
Posts: 719
Credit: 5,996,399
RAC: 2,455
Message 37635 - Posted: 18 Dec 2018, 11:28:37 UTC - in response to Message 37633.  

I am a bit more interested by your successful tasks, e.g.
https://lhcathome.cern.ch/lhcathome/result.php?resultid=212498881
https://lhcathome.cern.ch/lhcathome/result.php?resultid=212498882
https://lhcathome.cern.ch/lhcathome/result.php?resultid=212498883

They ran for quite a while (i.e. >1h), and I don't understand how it is possible that they did not hit the disk barrier...

I don't understand either, cause I digged into the result with id 212498881 and found the mem_alloc file a few minutes before the task finished.
I was monitoring big files for LHC-dev > 524,288 bytes.

17-Dec-2018 14:03:28 [LHC@home] Starting task sixtrack5p1p1_hl13B1__1__s__62.31_60.32__1_2__5__80_1_sixvf_boinc33_2
17-Dec-2018 15:45:05 [LHC@home] Computation for task sixtrack5p1p1_hl13B1__1__s__62.31_60.32__1_2__5__80_1_sixvf_boinc33_2 finished

Big files in slot 7 (slot for above task)
localtime ...... mod time .. size.......file
15:40:22.14 12:50:15 4517924 ".\7\fort.16"
15:40:22.14 15:38:52 728247 ".\7\fort.6"
15:40:22.14 14:03:28 4611986 ".\7\fort.9"

15:41:27.21 12:50:15 4517924 ".\7\fort.16"
15:41:27.21 14:03:28 1177848 ".\7\fort.31"
15:41:27.21 15:38:52 728247 ".\7\fort.6"
15:41:27.21 14:03:28 4611986 ".\7\fort.9"
15:41:27.21 15:41:16 739613414 ".\7\mem_alloc.log"

15:42:34.19 12:50:15 4517924 ".\7\fort.16"
15:42:34.19 14:03:28 1177848 ".\7\fort.31"
15:42:34.19 15:41:50 729597 ".\7\fort.6"
15:42:34.19 14:03:28 4611986 ".\7\fort.9"
15:42:34.19 15:41:16 739613414 ".\7\mem_alloc.log"

15:43:37.11 12:50:15 4517924 ".\7\fort.16"
15:43:37.11 14:03:28 1177848 ".\7\fort.31"
15:43:37.11 15:41:50 729597 ".\7\fort.6"
15:43:37.11 14:03:28 4611986 ".\7\fort.9"
15:43:37.11 15:41:16 739613414 ".\7\mem_alloc.log"

15:44:41.16 12:50:15 4517924 ".\7\fort.16"
15:44:41.16 14:03:28 1177848 ".\7\fort.31"
15:44:41.16 15:41:50 729597 ".\7\fort.6"
15:44:41.16 14:03:28 4611986 ".\7\fort.9"
15:44:41.16 15:41:16 739613414 ".\7\mem_alloc.log"

Next task in slot 7: 15:45:46.17 12:51:33 4517924 ".\7\fort.16"
ID: 37635 · Report as offensive     Reply Quote
Harri Liljeroos
Avatar

Send message
Joined: 28 Sep 04
Posts: 368
Credit: 19,578,534
RAC: 35,702
Message 37755 - Posted: 14 Jan 2019, 18:38:57 UTC

New tasks for sixtracktest 501.01 just arrived to two of my hosts (one for each). They are still in the queue waiting to be crunched.
ID: 37755 · Report as offensive     Reply Quote
Erich56

Send message
Joined: 18 Dec 15
Posts: 1119
Credit: 20,832,632
RAC: 37,124
Message 37756 - Posted: 14 Jan 2019, 19:02:05 UTC - in response to Message 37755.  

although 484 such tasks should be available according to the Project Status Page, my host doesn't download any :-(
ID: 37756 · Report as offensive     Reply Quote
Richie_unstable

Send message
Joined: 26 Oct 18
Posts: 32
Credit: 777,166
RAC: 2
Message 37757 - Posted: 14 Jan 2019, 21:09:16 UTC - in response to Message 37756.  

although 484 such tasks should be available according to the Project Status Page, my host doesn't download any :-(


An hour later I see theres over 10K tasks in progress. My understanding is that server status doesn't get updated very often here. I would guess all the tasks were already delivered even if server page still showed some tasks being available. 484 is moderately small portion of 10K and there must be plenty of hosts that all together will suck that amount instantly if it's available. Perhaps you computer was late.
ID: 37757 · Report as offensive     Reply Quote
Profile Veronica
Project developer
Project tester
Project scientist
Avatar

Send message
Joined: 29 Sep 17
Posts: 14
Credit: 4,188,056
RAC: 17,845
Message 37892 - Posted: 1 Feb 2019, 10:15:39 UTC - in response to Message 37635.  
Last modified: 1 Feb 2019, 10:16:01 UTC

I don't understand either, cause I digged into the result with id 212498881 and found the mem_alloc file a few minutes before the task finished.
I was monitoring big files for LHC-dev > 524,288 bytes.


Just an update: This was a debug file we used during development of SixTrack 5 when we added dynamic allocation of the large tracking arrays. This should not have gone into the BOINC executables and has been fixed in the current test execs.
SixTrack 5 Core Developer. github.com/SixTrack/SixTrack
ID: 37892 · Report as offensive     Reply Quote
Profile MAGIC Quantum Mechanic
Avatar

Send message
Joined: 24 Oct 04
Posts: 840
Credit: 36,847,513
RAC: 34,306
Message 37980 - Posted: 11 Feb 2019, 21:17:55 UTC - in response to Message 37892.  



Just an update: This was a debug file we used during development of SixTrack 5 when we added dynamic allocation of the large tracking arrays. This should not have gone into the BOINC executables and has been fixed in the current test execs.


Server status says Unsent sixtracktest 21986 for about 24 hours (or more) but they won't send out
ID: 37980 · Report as offensive     Reply Quote
Ola

Send message
Joined: 7 Apr 18
Posts: 20
Credit: 104,597
RAC: 511
Message 37992 - Posted: 12 Feb 2019, 12:44:36 UTC

There is over 124k sixtracktest tasks available and no in progress. What's wrong?
ID: 37992 · Report as offensive     Reply Quote
Alessio Mereghetti
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 29 Feb 16
Posts: 141
Credit: 1,744,986
RAC: 5,096
Message 37993 - Posted: 12 Feb 2019, 13:30:54 UTC - in response to Message 37992.  

we are inquiring the reason for that with the IT guys - will come back to you once the situation is clarified.
ID: 37993 · Report as offensive     Reply Quote
Profile Nils Høimyr
Volunteer moderator
Project administrator
Project developer
Project tester

Send message
Joined: 15 Jul 05
Posts: 207
Credit: 4,437,252
RAC: 742
Message 37994 - Posted: 12 Feb 2019, 14:18:16 UTC - in response to Message 37993.  

I wonder if there is something with the task deadline or similar that prevents tasks from being picked up. From the server side, Sixtracktest tasks are pushed to the scheduler shared memory. However, my BOINC client cannot get any task either.
ID: 37994 · Report as offensive     Reply Quote
Alessio Mereghetti
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 29 Feb 16
Posts: 141
Credit: 1,744,986
RAC: 5,096
Message 38002 - Posted: 13 Feb 2019, 11:02:49 UTC - in response to Message 37994.  
Last modified: 13 Feb 2019, 11:03:15 UTC

an update - there seems to be an issue with a feature recently added in the BOINC server code, that was used for submitting those tasks.
The tasks have been deleted, and we are debugging the feature - thanks Nils!
ID: 38002 · Report as offensive     Reply Quote
Harri Liljeroos
Avatar

Send message
Joined: 28 Sep 04
Posts: 368
Credit: 19,578,534
RAC: 35,702
Message 38054 - Posted: 22 Feb 2019, 19:00:44 UTC

SIXTRACKTEST tasks are available again but I seem to be unable to download them again.
ID: 38054 · Report as offensive     Reply Quote
mrchips

Send message
Joined: 16 May 14
Posts: 9
Credit: 3,305,795
RAC: 2,894
Message 38055 - Posted: 22 Feb 2019, 20:19:41 UTC

I can't download any either, why?
ID: 38055 · Report as offensive     Reply Quote
Profile MAGIC Quantum Mechanic
Avatar

Send message
Joined: 24 Oct 04
Posts: 840
Credit: 36,847,513
RAC: 34,306
Message 38057 - Posted: 23 Feb 2019, 10:04:39 UTC

2/23/2019 1:56:33 AM | LHC@home | No tasks are available for sixtracktest

sixtracktest 134332 Unsent 385 In progress
ID: 38057 · Report as offensive     Reply Quote
Toby Broom
Volunteer moderator

Send message
Joined: 27 Sep 08
Posts: 551
Credit: 335,338,196
RAC: 648,528
Message 38060 - Posted: 23 Feb 2019, 11:45:19 UTC

my mac has some so maybe they are all mac tasks?
ID: 38060 · Report as offensive     Reply Quote
Guiri-One[Andalucia]

Send message
Joined: 1 Feb 06
Posts: 43
Credit: 9,723
RAC: 0
Message 38361 - Posted: 21 Mar 2019, 11:50:35 UTC

Hi all,

Any update about ARM tasks for any of the applications?

My Pi will come in few days...
ID: 38361 · Report as offensive     Reply Quote
maeax

Send message
Joined: 2 May 07
Posts: 715
Credit: 26,438,414
RAC: 27,798
Message 39330 - Posted: 10 Jul 2019, 3:33:43 UTC

Are we on the way to virtualize the LHC?
Three Computer run 3 Days for this sixtracktest successful ;-)
https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=117930634
ID: 39330 · Report as offensive     Reply Quote
Alessio Mereghetti
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 29 Feb 16
Posts: 141
Credit: 1,744,986
RAC: 5,096
Message 39334 - Posted: 11 Jul 2019, 7:51:03 UTC - in response to Message 39330.  

Hi all,
Any update about ARM tasks for any of the applications?
My Pi will come in few days...


Hello Guiri-One[Andalucia],
sorry for the late reply. We do have both sixtrack and sixtracktest being distributed with ARM exes - we haven't solved all the issues with (no longer so) recent Android distributions. Anyway, I think your Pi should have already received and crunched some tasks...

Are we on the way to virtualize the LHC?
Three Computer run 3 Days for this sixtracktest successful ;-)
https://lhcathome.cern.ch/lhcathome/workunit.php?wuid=117930634


Indeed, with 10^7 turns, we can simulate 800s of beam dynamics. It is not the typical time we keep the beam in the LHC (~several hours), but we can simulate more than 50% of the energy ramp, for instance.
Anyway, these were just some tasks so that long - I am not sure they are efficient, and we are looking into segmenting the simulation steps - e.g. 10 jobs of 10^6 turns run in sequence instead of a big job with 10^7 turns.
For that, we have to make use of the check-point/restart option, and we have to send the necessary files as starting conditions and at the end of the simulation + checking that there are particles to be tracked.
We should soon be able to have this.

Thanks a lot for the support, keep up the good work, and happy crunching!
ID: 39334 · Report as offensive     Reply Quote
maeax

Send message
Joined: 2 May 07
Posts: 715
Credit: 26,438,414
RAC: 27,798
Message 39335 - Posted: 11 Jul 2019, 9:18:49 UTC - in response to Message 39334.  


Anyway, these were just some tasks so that long - I am not sure they are efficient, and we are looking into segmenting the simulation steps - e.g. 10 jobs of 10^6 turns run in sequence instead of a big job with 10^7 turns.
For that, we have to make use of the check-point/restart option, and we have to send the necessary files as starting conditions and at the end of the simulation + checking that there are particles to be tracked.
We should soon be able to have this.

Hi Alessio,
to split the longrunners is a good idea.
Atlas give us 200 Beams and this is a lot of CPU-Time to do this.
ID: 39335 · Report as offensive     Reply Quote
maeax

Send message
Joined: 2 May 07
Posts: 715
Credit: 26,438,414
RAC: 27,798
Message 39535 - Posted: 9 Aug 2019, 5:59:56 UTC

Have a Computer with only sixtracktest for the Moment:
The wingman had this Error:
exceeded elapsed time limit 13480.39 (1920000000.00G/109981.44G)</message>
https://lhcathome.cern.ch/lhcathome/result.php?resultid=238806725
ID: 39535 · Report as offensive     Reply Quote
Previous · 1 . . . 3 · 4 · 5 · 6 · 7 · 8 · Next

Message boards : Sixtrack Application : SIXTRACKTEST


©2019 CERN