[gmx-users] magic number error

[Erik Lindahl]

giacomo de mori wrote:
> Hello All
> some of my simulation suddenly  abort without any message in the log
> file.
> when I run gmxdump, at the last line it appears this message
> 
> in the first simulation:
> 
> Fatal error: Magic Number Error in XTC file (read -1546048969, should be
> 1995)
> 
> in the second simulation:
> Fatal error: Magic Number Error in XTC file (read -834587746, should be
> 1995)
> 
> similar message apperars in other simulations.
> Each simulation is a molecular dynamic of a 36 residue protein in
> explicit water using PME in gromacs 3.0.5.
> what does it mean?  How could I resolve this problem?
> thak you in advance

Did you fill your disk during the runs, or have you had a disk crash, 
system crash or something?

The magic number is just a special number entry at the beginning of each 
frame to guarantee that it is a Gromacs trajectory (in contrast to any 
other file with the .xtc extension).

The only way this could be wrong (in particular if it is right in the 
first part of the trajectory) is if your file has been corrupted.

Cheers,

Erik