Re: problem caused by cluster?

From: Jim Phillips (jim_at_ks.uiuc.edu)
Date: Fri Oct 08 2010 - 08:34:23 CDT

Hi,

The error is because the extraBonds feature is being used to add angles
for which some of the atoms are as much as 20A apart. All of the atoms in
every bond, angle, dihedral, or improper must be contained 2 x 2 x 2
blocks of patches. Otherwise on a large enough run a bad global X count
error will occur.

If you really want to do this, set twoAwayX, twoAwayY, and twoAwayZ all to
"no". If that doesn't fix it (in your specific case it should), increase
margin until your patches are large enough.

FYI, documentation for the extraBonds feature is now in CVS.

-Jim

On Fri, 8 Oct 2010, Kwee Hong wrote:

> Hi Jim,
>
> I've attached my input file to run this production stage in this email.
> Before this, I've done 1000 steps minimization, applying pressure for 6ps
> and 20ps equilibration.
>
> Regards,
> Joyce
>
> On Fri, Oct 8, 2010 at 4:41 AM, Jim Phillips <jim_at_ks.uiuc.edu> wrote:
>
>>
>> This kind of error gets hidden on serial runs. Can you send me a link to a
>> .tar.gz of all of the input files needed to reproduce this?
>>
>> -Jim
>>
>>
>>
>> On Wed, 6 Oct 2010, Kwee Hong wrote:
>>
>> Hi,
>>>
>>> I extend a 1ns simulation which was run on a single work station on a 14
>>> nodes cluster, But I got this error:
>>>
>>> *FATAL ERROR: Bad global angle count!*
>>> *
>>> *
>>> *FATAL ERROR: See http://www.ks.uiuc.edu/Research/namd/bugreport.html*
>>> *------------- Processor 0 Exiting: Called CmiAbort ------------*
>>> *Reason: FATAL ERROR: Bad global angle count!*
>>> *
>>> *
>>> *FATAL ERROR: See http://www.ks.uiuc.edu/Research/namd/bugreport.html*
>>>
>>> After reading the NAMD troubleshooting and some posts at the mailing list,
>>> I'm sure that i use the correct .xsc file at the restart with the margin
>>> set
>>> at 2.5 in the config file. And when i visualised my system using vmd, I do
>>> not find any long bonds. I don't understand why such situation happen.
>>>
>>> Then I tried to run the simulation from the beginning (not using any
>>> restart
>>> file), it return the same error, too. Is there any problem with my cluster
>>> as I had the same system run on a single workstation for 1ns and it
>>> completed successfully?
>>>
>>> Regards,
>>> Joyce
>>>
>>>
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:54:35 CST