[gmx-users] decomposition error
Chang Woon Jang
changwoonjang at gmail.com
Sat Jul 23 03:34:11 CEST 2016
Dear Mark Abraham,
I am a little confused about the number of ranks. How can I choose the
number of ranks? Is this mdrun option or something else? Is "number of
ranks" different with "number of processors" in parallel machine?
Thank you very much for your help.
Best regards,
Changwoon Jang
On Fri, Jul 22, 2016 at 9:16 PM, Mark Abraham <mark.j.abraham at gmail.com>
wrote:
> Hi,
>
> On Sat, 23 Jul 2016 03:05 Chang Woon Jang <changwoonjang at gmail.com> wrote:
>
> > Dear Mark Abraham,
> >
> > Thank you for the details. Do 27, 18, or 12 mean -dds option or
> > something else?
>
>
> No, the number of ranks. -dds pertains to how much margin DD requires to
> make sure the largest interactions will be computable over the lifetime of
> the decomposition as particles move. This comes back to my question about
> the ranges of your largest interactions, and how fast things move.
>
> How can I set up proper DD option?
> >
>
> DD will do what it can with the number of ranks you choose, but it can't
> handle factors of bigger than three for a system like yours. One way it can
> use 16 is in a 4x2x2 grid but it can't do that.
>
> Mark
>
>
> > Thank you for your advice.
> >
> > Best regards,
> > Changwoon Jang
> >
> > On Fri, Jul 22, 2016 at 8:58 PM, Mark Abraham <mark.j.abraham at gmail.com>
> > wrote:
> >
> > > Hi,
> > >
> > > Well, is it reasonable that your system has a dihedral where the
> furthest
> > > particles are 1.4nm apart? If so, then your system is quite small with
> > > respect to the range of interactions within it, and as that output
> notes
> > > mdrun is limited to three ranks in each direction. So 27, 18 or 12 can
> > give
> > > you more joy
> > >
> > > Mark
> > >
> > > On Sat, 23 Jul 2016 01:42 Chang Woon Jang <changwoonjang at gmail.com>
> > wrote:
> > >
> > > > Dear Justin Lemkul,
> > > >
> > > > I have looked at the log file as below.
> > > >
> > > > Initializing Domain Decomposition on 16 ranks
> > > >
> > > > Dynamic load balancing: auto
> > > >
> > > > Will sort the charge groups at every domain (re)decomposition
> > > >
> > > > Initial maximum inter charge-group distances:
> > > >
> > > > two-body bonded interactions: 0.668 nm, Tab. Bonds, atoms 1277
> 1278
> > > >
> > > > multi-body bonded interactions: 1.460 nm, Tab. Dih., atoms 1226
> 1229
> > > >
> > > > Minimum cell size due to bonded interactions: 1.500 nm
> > > >
> > > > Scaling the initial minimum size with 1/0.8 (option -dds) = 1.25
> > > >
> > > > Optimizing the DD grid for 16 cells with a minimum initial size of
> > 1.875
> > > nm
> > > >
> > > > The maximum allowed number of cells is: X 3 Y 3 Z 3
> > > >
> > > >
> > > > Does this mean that I need to use "-dds 1.25" mdrun option?
> > > >
> > > > Thank you.
> > > >
> > > > Best regards,
> > > > Changwoon Jang
> > > >
> > > > On Fri, Jul 22, 2016 at 7:07 PM, Justin Lemkul <jalemkul at vt.edu>
> > wrote:
> > > >
> > > > >
> > > > >
> > > > > On 7/22/16 7:03 PM, Chang Woon Jang wrote:
> > > > >
> > > > >> Dear Justin A. Lumkul,
> > > > >>
> > > > >> Thank you for your answer. I am sorry for asking what is DD. My
> > > > >> topology
> > > > >> topol.top file is below. Would you please tell me about what is
> DD?
> > > How
> > > > >> can
> > > > >> I set up in topology file?
> > > > >>
> > > > >>
> > > > > DD = domain decomposition. Please look in the .log file for how it
> > is
> > > > > being set up and provide that information if you need help
> > > > troubleshooting
> > > > > it. Also use Google; this error comes up (and gets solved) every
> > > couple
> > > > of
> > > > > days.
> > > > >
> > > > > -Justin
> > > > >
> > > > >
> > > > > Thank you.
> > > > >>
> > > > >> Best regards,
> > > > >> Changwoon Jang
> > > > >>
> > > > >>
> > > > >>
> > > > >> [ defaults ]
> > > > >>
> > > > >> ; nbfunc comb-rule gen-pairs fudgeLJ fudgeQQ
> > > > >>
> > > > >> 1 1 no 0.0 0.0
> > > > >>
> > > > >>
> > > > >> [ atomtypes ]
> > > > >>
> > > > >> ;type mass charge ptype sigma epsilon
> > > > >>
> > > > >> A 75.087799 0.000 A 1.000000000 1.000000000
> > > > >>
> > > > >> B 76.097998 0.000 A 1.000000000 1.000000000
> > > > >>
> > > > >> C 42.080999 0.000 A 1.000000000 1.000000000
> > > > >>
> > > > >> D 74.103099 0.000 A 1.000000000 1.000000000
> > > > >>
> > > > >> E 99.153398 0.000 A 1.000000000 1.000000000
> > > > >>
> > > > >> F 116.160799 0.000 A 1.000000000 1.000000000
> > > > >>
> > > > >> G 58.080399 0.000 A 1.000000000 1.000000000
> > > > >>
> > > > >>
> > > > >> [ moleculetype ]
> > > > >>
> > > > >> DGA 3
> > > > >>
> > > > >>
> > > > >> [atoms]
> > > > >>
> > > > >> ; nr type resnr residue atom cgnr charge mass
> > > > >>
> > > > >> 1 A 1 DGA A1 1 0.000000 75.087799
> > > > >>
> > > > >> 2 B 1 DGA B1 2 0.000000 76.097998
> > > > >>
> > > > >> 3 C 1 DGA C 3 0.000000 42.080999
> > > > >>
> > > > >> 4 B 1 DGA B2 4 0.000000 76.097998
> > > > >>
> > > > >> 5 A 1 DGA A2 5 0.000000 75.087799
> > > > >>
> > > > >>
> > > > >>
> > > > >> [ bonds ]
> > > > >>
> > > > >> 1 2 8 1 1.0 ; 1:bond:1
> > > > >>
> > > > >> 2 3 8 1 1.0 ; 1:bond:2
> > > > >>
> > > > >> 3 4 8 1 1.0 ; 1:bond:3
> > > > >>
> > > > >> 4 5 8 1 1.0 ; 1:bond:4
> > > > >>
> > > > >>
> > > > >> [ angles ]
> > > > >>
> > > > >> 1 2 3 8 1 1.0 ; 1:angle:1
> > > > >>
> > > > >> 2 3 4 8 1 1.0 ; 1:angle:2
> > > > >>
> > > > >> 3 4 5 8 1 1.0 ; 1:angle:3
> > > > >>
> > > > >>
> > > > >> [ dihedrals ]
> > > > >>
> > > > >> 1 2 3 4 8 1 1.0 ; 1:ABCBdihedral:1
> > > > >>
> > > > >> 2 3 4 5 8 1 1.0 ; 1:ABCBdihedral:2
> > > > >>
> > > > >>
> > > > >> [ moleculetype ]
> > > > >>
> > > > >> J400 3
> > > > >>
> > > > >>
> > > > >> [atoms]
> > > > >>
> > > > >> ; nr type resnr residue atom cgnr charge mass
> > > > >>
> > > > >> 1 D 1 J400 D1 1 0.000000 74.103099
> > > > >>
> > > > >> 2 E 1 J400 E1 2 0.000000 99.153398
> > > > >>
> > > > >> 3 F 1 J400 F1 3 0.000000 116.160799
> > > > >>
> > > > >> 4 G 1 J400 G1 4 0.000000 58.080399
> > > > >>
> > > > >> 5 D 1 J400 D2 5 0.000000 74.103099
> > > > >>
> > > > >>
> > > > >> [ bonds ]
> > > > >>
> > > > >> 1 2 8 1 1.0 ; 1:bond:1
> > > > >>
> > > > >> 2 3 8 1 1.0 ; 1:bond:2
> > > > >>
> > > > >> 3 4 8 1 1.0 ; 1:bond:3
> > > > >>
> > > > >> 4 5 8 1 1.0 ; 1:bond:4
> > > > >>
> > > > >>
> > > > >> [ angles ]
> > > > >>
> > > > >> 1 2 3 8 1 1.0 ; 1:angle:1
> > > > >>
> > > > >> 2 3 4 8 1 1.0 ; 1:angle:2
> > > > >>
> > > > >> 3 4 5 8 1 1.0 ; 1:angle:3
> > > > >>
> > > > >>
> > > > >> [ dihedrals ]
> > > > >>
> > > > >> 1 2 3 4 8 1 1.0 ; 1:dihedral:1
> > > > >>
> > > > >> 2 3 4 5 8 1 1.0 ; 1:dihedral:2
> > > > >>
> > > > >>
> > > > >> [ system ]
> > > > >>
> > > > >> ; Name
> > > > >>
> > > > >> Built with Packmol
> > > > >>
> > > > >>
> > > > >> [ molecules ]
> > > > >>
> > > > >> ; Compound #mols
> > > > >>
> > > > >> DGA 200
> > > > >>
> > > > >> J400 100
> > > > >>
> > > > >>
> > > > >> On Fri, Jul 22, 2016 at 6:59 PM, Justin Lemkul <jalemkul at vt.edu>
> > > wrote:
> > > > >>
> > > > >>
> > > > >>>
> > > > >>> On 7/22/16 6:50 PM, Chang Woon Jang wrote:
> > > > >>>
> > > > >>> Dear Grimaces Users,
> > > > >>>>
> > > > >>>> I have the decomposition error. I have tried to reduce the
> number
> > of
> > > > >>>> cores
> > > > >>>> but still have the same problem. In my conf.gro, the cell size
> is
> > > > about
> > > > >>>> 5.5x5.5x5.5 but I do not know what the minimum cell size of
> > 1.875nm.
> > > > >>>>
> > > > >>>> Would you please give me some comments or relevant topics
> > > addressed?
> > > > >>>>
> > > > >>>>
> > > > >>>> The .log file tells you how DD is being set up and what the the
> > > > limiting
> > > > >>> interactions are. Verify that your topology is sound from what
> the
> > > > .log
> > > > >>> tells you.
> > > > >>>
> > > > >>> -------------------------------------------------------
> > > > >>>
> > > > >>>>
> > > > >>>> Program gmx, VERSION 5.0.7-dev-20151003-1909f2f
> > > > >>>>
> > > > >>>> Source code file:
> > > > >>>>
> > > > >>>>
> > > >
> > >
> >
> /home1/02271/huddack/votca_install/src/gromacs/src/gromacs/mdlib/domdec.c,
> > > > >>>> line: 6902
> > > > >>>>
> > > > >>>>
> > > > >>>> Also make sure your run isn't failing because of something due
> to
> > > > hacked
> > > > >>> code. Development versions generally shouldn't be used for
> > > > >>> production-level
> > > > >>> science.
> > > > >>>
> > > > >>> -Justin
> > > > >>>
> > > > >>>
> > > > >>> Fatal error:
> > > > >>>>
> > > > >>>> There is no domain decomposition for 16 ranks that is compatible
> > > with
> > > > >>>> the
> > > > >>>> given box and a minimum cell size of 1.875 nm
> > > > >>>>
> > > > >>>> Change the number of ranks or mdrun option -rdd or -dds
> > > > >>>>
> > > > >>>> Look in the log file for details on the domain decomposition
> > > > >>>>
> > > > >>>> For more information and tips for troubleshooting, please check
> > the
> > > > >>>> GROMACS
> > > > >>>>
> > > > >>>> website at http://www.gromacs.org/Documentation/Errors
> > > > >>>>
> > > > >>>> -------------------------------------------------------
> > > > >>>>
> > > > >>>>
> > > > >>>> --
> > > > >>> ==================================================
> > > > >>>
> > > > >>> Justin A. Lemkul, Ph.D.
> > > > >>> Ruth L. Kirschstein NRSA Postdoctoral Fellow
> > > > >>>
> > > > >>> Department of Pharmaceutical Sciences
> > > > >>> School of Pharmacy
> > > > >>> Health Sciences Facility II, Room 629
> > > > >>> University of Maryland, Baltimore
> > > > >>> 20 Penn St.
> > > > >>> Baltimore, MD 21201
> > > > >>>
> > > > >>> jalemkul at outerbanks.umaryland.edu | (410) 706-7441
> > > > >>> http://mackerell.umaryland.edu/~jalemkul
> > > > >>>
> > > > >>> ==================================================
> > > > >>> --
> > > > >>> Gromacs Users mailing list
> > > > >>>
> > > > >>> * Please search the archive at
> > > > >>> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List
> before
> > > > >>> posting!
> > > > >>>
> > > > >>> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > > > >>>
> > > > >>> * For (un)subscribe requests visit
> > > > >>>
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users
> > > or
> > > > >>> send a mail to gmx-users-request at gromacs.org.
> > > > >>>
> > > > >>>
> > > > > --
> > > > > ==================================================
> > > > >
> > > > > Justin A. Lemkul, Ph.D.
> > > > > Ruth L. Kirschstein NRSA Postdoctoral Fellow
> > > > >
> > > > > Department of Pharmaceutical Sciences
> > > > > School of Pharmacy
> > > > > Health Sciences Facility II, Room 629
> > > > > University of Maryland, Baltimore
> > > > > 20 Penn St.
> > > > > Baltimore, MD 21201
> > > > >
> > > > > jalemkul at outerbanks.umaryland.edu | (410) 706-7441
> > > > > http://mackerell.umaryland.edu/~jalemkul
> > > > >
> > > > > ==================================================
> > > > > --
> > > > > Gromacs Users mailing list
> > > > >
> > > > > * Please search the archive at
> > > > > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > > > > posting!
> > > > >
> > > > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > > > >
> > > > > * For (un)subscribe requests visit
> > > > > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users
> > or
> > > > > send a mail to gmx-users-request at gromacs.org.
> > > > >
> > > >
> > > >
> > > >
> > > > --
> > > > Best regards,
> > > > Changwoon Jang,
> > > >
> > > > Postdoctoral Research Fellow
> > > > Department of Chemical & Biological Engineering, Drexel University
> > > > 3141 Chestnut Street, Philadelphia, PA 19104
> > > >
> > > > Voice: (662) 617-2267
> > > > E-mail: cj395 at drexel.edu
> > > > --
> > > > Gromacs Users mailing list
> > > >
> > > > * Please search the archive at
> > > > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > > > posting!
> > > >
> > > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > > >
> > > > * For (un)subscribe requests visit
> > > > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users
> or
> > > > send a mail to gmx-users-request at gromacs.org.
> > > >
> > > --
> > > Gromacs Users mailing list
> > >
> > > * Please search the archive at
> > > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > > posting!
> > >
> > > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> > >
> > > * For (un)subscribe requests visit
> > > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> > > send a mail to gmx-users-request at gromacs.org.
> > >
> > --
> > Gromacs Users mailing list
> >
> > * Please search the archive at
> > http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> > posting!
> >
> > * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
> >
> > * For (un)subscribe requests visit
> > https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> > send a mail to gmx-users-request at gromacs.org.
> >
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-request at gromacs.org.
>
More information about the gromacs.org_gmx-users
mailing list