[gmx-users] defining parameters in a separate itp file

[Justin Lemkul]

On 11/12/19 6:30 PM, Dave M wrote:
> Hi Justin,
>
> I am having troubles now with domain decompositions when I use openMPI
> mpirun but it works fine when I run on single mpi. My system is quite large
> (80x80x80) and I defined additional bonds in top file as:
>
>
> [ intermolecular_interactions ]
>
> [ bonds ]
>
> ; ai aj type
>
> 3123   40767   6   4.0047  1000
>
> 812    41625   6   3.7768  1000
>
> 1560   42483   6   3.7569  1000
>
> 2378   43341   6   4.0945  1000
>
>
>
> I get error like this:
>
>
> Not all bonded interactions have been properly assigned to the domain
> decomposition cells
>
> A list of missing interactions:
>
>         Harmonic Pot. of      8 missing      6
>
>            exclusions of     -4 missing     -4
>
>
> I saw an old mail thread where Mark suggests to use
> [intermolecular_interactions]. Am already doing it but I still get errors
> (link:
> https://www.mail-archive.com/gromacs.org_gmx-users@maillist.sys.kth.se/msg37888.html
>   )
>
>
> I use this:
>
> mpirun -np 4 gmx_mpi mdrun  -deffnm 03-run -rdd 2 -ntomp 16
>
>
> With np = 1 it works fine.
>
>
> Would it be possible to overcome this? On single MPI it will take many
> months to get the job done.

Everything you've posted seems to suggest that the way 
[intermolecular_interactions] are designed to function no longer work 
properly. A bug report/feature request may be appropriate via Redmine.

In the meantime, it looks like your only option is to use the pull code, 
which does play nicely with domain decomposition.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Assistant Professor
Office: 301 Fralin Hall
Lab: 303 Engel Hall

Virginia Tech Department of Biochemistry
340 West Campus Dr.
Blacksburg, VA 24061

jalemkul at vt.edu | (540) 231-3129
http://www.thelemkullab.com

==================================================