[gmx-users] FAQ: File too big

[Oliver Beckstein]

Hi,

I wondered if anyone has finally found out to how fix the annoying
'File too big problem' (the latest post was
http://www.gromacs.org/pipermail/gmx-users/2003-April/031896.html
). David suggests the workaround to access the file through nfs (which
works for me) but it is not really the thing that I want to do with a
few times 3+ GB of data.

Splitting the xtc is not a good solution as I need a continous time
series (perhaps one could allow the analysis tools to read multiple
xtc's which they concat internally?).

For the archive:

The problem manifests itself as the message 'File not found' when
any of the analysis tools (including gmxcheck) refeuses to read a file
which is clearly there (and can be manipulated with standard unix
tools (cat, cp, split. hexdump, ...).

In detail:

ls -l big.xtc:
-rw-r--r--    1 oliver oliver 2735936052 May 21 17:48 big.xtc

strace gmxcheck -f big.xtc:
[...]
open("big.xtc", O_RDONLY) = -1 EFBIG (File too large)


Howvere, I was under the impression that my kernel Linux
2.4.18-27.8.0smp (RedHat vanilla) has full large file support, and
configure's output looks promising:

[...]
checking for special C compiler options needed for large files... no
checking for _FILE_OFFSET_BITS value needed for large files... 64
checking for _LARGE_FILES value needed for large files... no
checking for _LARGEFILE_SOURCE value needed for large files... 1
[...]

Perhaps someone out there knows a simple solution?

Oli


-- 
Oliver Beckstein * oliver at bioch.ox.ac.uk
 http://indigo1.biop.ox.ac.uk/oliver/