RE: problems reading large dcd file

From: Joshua D. Moore (jdmoore_at_unity.ncsu.edu)
Date: Tue Jan 10 2006 - 19:35:55 CST

Hi Jorge and everyone,

There have been some recent additions to the AIX Wiki

http://www.ks.uiuc.edu/Research/namd/wiki/index.cgi?NamdOnAIX

Some of the comments relate to large file support. Jorge, are you running
on an AIX machine?

Josh

On Tue, January 10, 2006 11:59 am, Michel Espinoza-Fonseca wrote:
> Jorge,
>
> Regarding the first question, you shouldn't have any problem at loading
> dcd files in either 32 or 64-bit machines. For example, I usually run my
> simulations using an Altix cluster (64 bits) and perform the analysis in a
> PC running under Linux or Windows (32 bits). I don't know exactly what's
> your problem, but I've been using catDCD to cut very big dcd files and I
> haven't encounter any problem.
> Second, what do you mean with "recover" a part of the trajectory?
>
> Mucha suerte,
> Michel
>
> ________________________________
>
> From: owner-namd-l_at_ks.uiuc.edu on behalf of Jorge Pikunic
> Sent: Tue 1/10/2006 10:20 AM
> To: jdmoore_at_unity.ncsu.edu
> Cc: namd-l_at_ks.uiuc.edu
> Subject: RE: namd-l: problems reading large dcd file
>
>
>
> Hi Josh,
>
> Thank you and others for replying. It seems that the version of NAMD that
> I
> was using was compiled in 32 bit. The dcd files are 2.2GB. I have the
> following questions (maybe the developers can help answering these):
> 1. Is it likely that the dcd files have been corrupted due to the 32 bit
> addressing?
> 2. Is there any way of recovering at least part of the trajectory stored
> in
> those dcd files?
>
> Thanks again,
>
> Jorge
>
>
> Dr. Jorge Pikunic
> Department of Biochemistry
> University of Oxford
> South Parks Road
> Oxford OX1 3QU
> United Kingdom
>
>
>> -----Original Message-----
>> From: Joshua D. Moore [mailto:jdmoore_at_unity.ncsu.edu]
>> Sent: 07 January 2006 5:23 AM
>> To: jorge.pikunic_at_bioch.ox.ac.uk
>> Cc: namd-l_at_ks.uiuc.edu
>> Subject: Re: namd-l: problems reading large dcd file
>>
>> Hey Jorge,
>>
>> I am not sure if you are having the same problem as me. You say
>> different
>> system, do you mean different computer system? Anyway, my problem was
>> that the files were reaching their limit when I was running in 32 bit
>> mode
>> on DataStar. This made the file size limit 2 GB, and the DCD files were
>> unreadable (by VMD and catdcd). I am now recompiling NAMD in 64 bit
>> mode
>> and the people at SDSC say my problem should go away as there is no file
>> size limit.
>>
>> I was having a similar problem with catdcd 4.0, but I think that my
>> binary
>> may have been compiled in 32 bit. I used an old version of catdcd 3.0
>> before and I could combine files into larger than 2 GB so maybe it was
>> 64
>> bit (not really sure).
>>
>> Let me know if this is your problem or you find a solution.
>>
>> Josh
>>
>>
>> On Fri, January 6, 2006 11:45 am, Jorge Pikunic wrote:
>> > Hello,
>> >
>> > I am having problems reading dcd files from simulations of a fairly
>> > large system (200,000 atoms). The size of each file is about 2.1 GB. I
>> > have tried doing:
>> >
>> > 1. catdcd: catdcd -num production_1.dcd , and I get:
>> > CatDCD 3.0
>> > read_dcdreader returned -6
>> > Error: could not open file 'production_1.dcd' for reading.
>> >
>> > 2. vmd: When I try to load the dcd file, after loading the psf file, I
>> > get "unable to load molecule". I get the same result even if I try to
>> > load only one frame.
>> >
>> > I have been able to read larger files from different systems (> 5 GB)
>> > with catdcd. Also, I can read smaller dcd files (1.7 GB) from shorter
>> > simulations of the same system. I have checked the NAMD output files,
>> > but there is nothing strange; NAMD wrote to dcd files periodically and
>> > there are no errors reported.
>> >
>> > Any ideas of what might be wrong?
>> >
>> > Thank you,
>> >
>> > Jorge
>> >
>> >
>> >
>>
>>
>> --------------------------------------------------
>> Joshua D. Moore
>> Graduate Student
>> North Carolina State University
>> Department of Chemical and Biomolecular Engineering
>> Box 7905 Centennial Campus
>> Engineering Building I
>> 911 Partners Way
>> Raleigh, NC 27695-7905
>> Phone: (919) 513-2051
>> Fax: (919) 513-2470
>> Email: jdmoore_at_unity.ncsu.edu
>> --------------------------------------------------
>>
>
>
>
>
>
>

--------------------------------------------------
Joshua D. Moore
Graduate Student
North Carolina State University
Department of Chemical and Biomolecular Engineering
Box 7905 Centennial Campus
Engineering Building I
911 Partners Way
Raleigh, NC 27695-7905
Phone: (919) 513-2051
Fax: (919) 513-2470
Email: jdmoore_at_unity.ncsu.edu
--------------------------------------------------

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:41:31 CST