[gmx-users] Shake block errors with d.lzm

[David van der Spoel]

Ganesh CN wrote:
> Hi,
> 
> I am novice GROMACS user and I am trying to use the d.lzm benchmark of 
> GROMACS for a parallel MPI run.  Its working fine for the 8 processor 
> case, but from 16 processors onwards (tried until 256 processors), 
> running grompp is giving me errors for "Shake block crossing node 
> boundaries".  ( I am not a MD person at all, so I have no idea what this 
> means.).  The version I am using is 3.3.1 and the command line I used is
> 
> $ grompp -f cutoff.mdp -np 16
> 
> The exact error I am getting is :
> 
> Checking consistency between energy and charge groups...
> splitting topology...
> Walking down the molecule graph to make shake-blocks
> There are 7882 charge group borders and 7846 shake borders
> There are 7429 total borders
> Division over nodes in atoms:
>     1453    1451    1449    1449    1449    1452    1449    1452    
> 1449    1452    1449    1452    1449    1452    1449    1451
> -------------------------------------------------------
> Program grompp, VERSION 3.3.1
> Source code file: splitter.c, line: 121
> 
> Fatal error:
> Shake block crossing node boundaries
> constraint between atoms (1447,1454)
> -------------------------------------------------------
> 
> I saw that there were a couple of other threads which talked about this:
> 
> http://www.gromacs.org/pipermail/gmx-users/2005-January/013680.html
> http://www.gromacs.org/pipermail/gmx-developers/2004-January/000733.html 
> <http://www.gromacs.org/pipermail/gmx-developers/2004-January/000733.html>
> 
> But none of them had any solutions provided.  The second one said I 
> tried looking up the bugzilla and documentation for any bugs and found 
> this : http://bugzilla.gromacs.org/show_bug.cgi?id=110 
> <http://bugzilla.gromacs.org/show_bug.cgi?id=110>.  It seems to suggest 
> that there is something wrong with the topology file, but I am using the 
> standard topology file shipped with d.lzm.
> 
> Does anybody faced this problem before and has a solution ? Or is this a 
> known issue ?

you can not use so many processors for such a small system in 3.3. In 
4.0 you will be able to do it.

> 
> Thanks in advance,
> Ganesh
> 
> 
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> gmx-users mailing list    gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please search the archive at http://www.gromacs.org/search before posting!
> Please don't post (un)subscribe requests to the list. Use the 
> www interface or send it to gmx-users-request at gromacs.org.
> Can't post? Read http://www.gromacs.org/mailing_lists/users.php


-- 
David van der Spoel, Ph.D.
Molec. Biophys. group, Dept. of Cell & Molec. Biol., Uppsala University.
Box 596, 75124 Uppsala, Sweden. Phone:	+46184714205. Fax: +4618511755.
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://folding.bmc.uu.se