Re: FATAL ERROR: Due to a design error, GlobalMasterServer does not support individual atom requests from multiple global force clients on parallel runs.

From: Jérôme Hénin (jhenin_at_ifr88.cnrs-mrs.fr)
Date: Thu Sep 09 2010 - 11:06:33 CDT

Hao:

Good news: I have found the bug that caused your problem. Attached is
a source file containing the correction. I suppose you are running on
a supercomputer: I hope that you have a way to recompile NAMD there
(if needed, ask the system administrators for assistance). If you want
to run tests on a desktop machine and cannot recompile there, I can
try to provide you with a binary, if you tell me the architecture.

Note that you will have to choose between changing the target RMSD
(targetCenters) and changing the force constant (targetforceConstant).
Other than that, your input should work fine.

Thanks for reporting this bug!

Best wishes,
Jerome

On 9 September 2010 17:01, dhacademic <dhacademic_at_gmail.com> wrote:
> Attached please find two files:
>
> "colvarm.pdb" is used for "seg A A B B C C D D" and "(resid 386 to 485),
> (resid 628 to 673)", which contains 944 atoms;
> "colvarn.pdb" is used for "segA and (resid 386 to 485)", which contains 100
> atoms.
>
> Many thanks,
>
> Hao
>
>
>
> On Thu, Sep 9, 2010 at 10:31 AM, Jérôme Hénin <jhenin_at_ifr88.cnrs-mrs.fr>
> wrote:
>>
>> Hi,
>> We also need colvarn.pdb.
>> Jerome
>>
>> On 9 September 2010 00:27, dhacademic <dhacademic_at_gmail.com> wrote:
>> > Sorry, I forget to attach the parameter file in last email.
>> >
>> >
>> >
>> > On Wed, Sep 8, 2010 at 6:19 PM, dhacademic <dhacademic_at_gmail.com> wrote:
>> >>
>> >> Hi Giacomo and Jerome,
>> >>
>> >> Attached please find my files. Plese use "tar -xvzf my-namd.tar" to
>> >> unzip
>> >> the files.
>> >>
>> >> psf file: final-Glu-0630.psf
>> >> restart-coor file: myrst7a11.coor
>> >> namd conf file: dyn7a12.conf
>> >> colvars conf file: colvars.conf
>> >>
>> >> Many thanks,
>> >>
>> >> Hao
>> >>
>> >>
>> >>
>> >> On Wed, Sep 8, 2010 at 5:28 PM, Giacomo Fiorin
>> >> <giacomo.fiorin_at_temple.edu>
>> >> wrote:
>> >>>
>> >>> OK: I assumed that since you acknowledged that 2.7b1 was buggy, you
>> >>> were
>> >>> using 2.7b2 now, which should work.  If it still doesn't with 2.7b2,
>> >>> there
>> >>> may be a mismatch with the topology, or some bug we don't know of.
>> >>>
>> >>> What about trying the following in VMD?
>> >>>
>> >>> [atomselect top "(segid A) and (resid 386 to 485)"] get serial
>> >>> [atomselect top "(segid A) and (resid 628 to 673)"] get serial
>> >>> ... (and similar ones for B, C, and D) ...
>> >>>
>> >>> If the atom numbers you get are less than expected, then there is
>> >>> something wrong with the topology.  Otherwise, you should send to me
>> >>> and
>> >>> Jérôme the files, so that we investigate the error.  In the meantime,
>> >>> you
>> >>> can use the numbers provided by those VMD commands as an argument to
>> >>> atomNumbers in NAMD.  Does that work for you?
>> >>>
>> >>> Giacomo
>> >>>
>> >>>
>> >>> ---- ----
>> >>>   Dr. Giacomo Fiorin
>> >>>   ICMS - Institute for Computational Molecular Science - Temple
>> >>> University
>> >>>   1900 N 12 th Street, Philadelphia, PA 19122
>> >>>   giacomo.fiorin_at_temple.edu
>> >>> ---- ----
>> >>>
>> >>>
>> >>>
>> >>> On Tue, Sep 7, 2010 at 4:31 PM, dhacademic <dhacademic_at_gmail.com>
>> >>> wrote:
>> >>>>
>> >>>> Actually I had tried different combinations in atom definition part--005045029bef4ee84d048fd5d169--

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:56:07 CST