[gmx-users] REMD in one nodes with two processors

[Dongsheng Zhang]

David,

Thank you. My mdrun_mpi has problems with parallel computing. I don't
know why. I recompile gromacs, then it works now. I have saved the old
Makefile before "make distclean". I find it is the same as the new
Makefile, then I don't understand why the old mdrun_mpi didn't work.



Thank you very much for your help!


Dongsheng


On Thu, 2006-05-04 at 21:52 +0200, David van der Spoel wrote:
> Dongsheng Zhang wrote:
> > Thank you, David. I try it as:
> > 
> > 
> > mpirun -np 2  n1 /home/brasilia/dong/mdrun_mpi  -replex 500 -reseed
> > -1 -s replica -v >& zevel 
> > 
> > It gives me the same error message: Can not open file: replica.tpr
> > 
> was mdrun compiled with MPI? are you sure the second copy of mdrun is in 
> the right directory? try it from your home directory
> 
> > Any other suggestions?
> > 
> > Thanks!
> > 
> > 
> > Dongsheng
> > 
> > 
> > On Thu, 2006-05-04 at 20:17 +0200, David van der Spoel wrote:
> >> Dongsheng Zhang wrote:
> >>> Dear David,
> >>>
> >>> I had tried mpirun before, it didn't work either. the command was:
> >>> mpirun -np 2 n1 /home/brasilia/dong/mdrun_mpi_MC  -replex 500 -reseed -1
> >>> -multi -s replica -deffnm replica -N 2 -v >& zevel
> >> remove the -N 2 -multi -deffnm replica options and try again.
> >>
> >>
> >>>  Because the two processors are in one node, so I changed my command to
> >>> ( and go to the local node n1 to submit the job):
> >>>
> >>> /home/brasilia/dong/mdrun_mpi -np 2 -replex 500 -reseed -1 -multi -s
> >>> replica -deffnm replica -N 2 -v >& zevel
> >>>
> >>> Could you please tell me if there is any way to check if the program can
> >>> recognize -multi option?
> >>>
> >>> Thank you!
> >>>
> >>>
> >>> Dongsheng
> >>>
> >>>
> >>>
> >>> On Thu, 2006-05-04 at 19:13 +0200, David van der Spoel wrote:
> >>>> Dongsheng Zhang wrote:
> >>>>> Marcus,
> >>>>>
> >>>>> Thank you for your suggestion. I try it. The computer gives me the same
> >>>>> error message. The problem is that the program always looks for
> >>>>> replica.tpr instead of replica0.tpr and replica1.tpr. I don't know why.
> >>>>>
> >>>> use mpirun
> >>>> search mailinglist
> >>>>> Best wishes!
> >>>>>
> >>>>>
> >>>>> Dongsheng
> >>>>>
> >>>>>
> >>>>> On Thu, 2006-05-04 at 18:38 +0200, Marcus Kubitzki wrote:
> >>>>>> If you use the -deffnm option, you do not need -s, b/c it is
> >>>>>> already accounted for with the -deffnm.
> >>>>>> Marcus
> >>>>>>
> >>>>>> Dongsheng Zhang wrote:
> >>>>>>> David,
> >>>>>>>
> >>>>>>> Thank you for your reply. I have provided replica0.tpr and replica1.tpr,
> >>>>>>> but the program looks for replica.tpr. It surprises me.
> >>>>>>>
> >>>>>>> I have also tried to run 
> >>>>>>>
> >>>>>>> /home/brasilia/dong/mdrun_mpi -np 2 -replex 500 -reseed -1 -multi -s
> >>>>>>> replica0  replica1 -deffnm replica -N 2 -v >& zevel 
> >>>>>>>  It doesn't work either (complains only one replica: Nothing to exchange
> >>>>>>> with only one replica). it seems the program does not recognize the
> >>>>>>> -multi option. 
> >>>>>>>
> >>>>>>>
> >>>>>>> Best wishes!
> >>>>>>>
> >>>>>>>
> >>>>>>> Dongsheng
> >>>>>>>
> >>>>>>>
> >>>>>>> On Thu, 2006-05-04 at 18:11 +0200, David van der Spoel wrote:
> >>>>>>>
> >>>>>>>> Dongsheng Zhang wrote:
> >>>>>>>>
> >>>>>>>>> Dear gmx users,
> >>>>>>>>>
> >>>>>>>>> I have trouble to run REMD on one node with two processors. The command
> >>>>>>>>> I used is
> >>>>>>>>> /home/brasilia/dong/mdrun_mpi -np 2 -replex 500 -reseed -1 -multi -s
> >>>>>>>>> replica -deffnm replica -N 2 -v >& zevel
> >>>>>>>>>
> >>>>>>>>> the error message is:
> >>>>>>>>>
> >>>>>>>>> Can not open file:
> >>>>>>>>> replica.tpr
> >>>>>>>> you need replica0 and replica1.tpr
> >>>>>>>>
> >>>>>>>>> This error message looks strange to me. Because I have used -multi, the
> >>>>>>>>> program should look for replica0.tpr and replica1.tpr. Could anyone
> >>>>>>>>> point out what's wrong in my command?
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> Thank you in advance!
> >>>>>>>>>
> >>>>>>>>>
> >>>>>>>>> Dongsheng
> >>>>>>>>> _______________________________________________
> >>>>>>>>> gmx-users mailing list    gmx-users at gromacs.org
> >>>>>>>>> http://www.gromacs.org/mailman/listinfo/gmx-users
> >>>>>>>>> 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
> >>>>>>> _______________________________________________
> >>>>>>> gmx-users mailing list    gmx-users at gromacs.org
> >>>>>>> http://www.gromacs.org/mailman/listinfo/gmx-users
> >>>>>>> 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
> >>>>> _______________________________________________
> >>>>> gmx-users mailing list    gmx-users at gromacs.org
> >>>>> http://www.gromacs.org/mailman/listinfo/gmx-users
> >>>>> 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
> >>> _______________________________________________
> >>> gmx-users mailing list    gmx-users at gromacs.org
> >>> http://www.gromacs.org/mailman/listinfo/gmx-users
> >>> 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
> >>
> > _______________________________________________
> > gmx-users mailing list    gmx-users at gromacs.org
> > http://www.gromacs.org/mailman/listinfo/gmx-users
> > 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
> 
>