something wrong after reading the restart file

Dear all,
i saw in the manual that a restart file stores the following information about a simulation such as units and atom style, simulation box size shape and boundary settings, group definitions, per-type atom settings such as mass, per-atom attributes including their group assignments and molecular topology attributes, force field styles and coefficients, and special_bonds settings.
However, in the original input file, i used the “s s m” boundary settings in x, y, z directions, respectively. and the box size in z direction is lz=3.4 A. then in this input file, a restart file was output through the command of write_restart.
unfortunately, when i used the output restart file through the command of read_restart in another input file, i found that the box size in z direction had been changed to lz=0.000068 A. so i thought the boundary settings had also been changed to s s s in x, y, z directions, respectively.
is that a bug in lammps ?
thank you so much.

Dear all,
         i saw in the manual that a restart file stores the following
information about a simulation such as units and atom style, simulation box
size shape and boundary settings, group definitions, per-type atom settings
such as mass, per-atom attributes including their group assignments and
molecular topology attributes, force field styles and coefficients, and
special_bonds settings.
        However, in the original input file, i used the "s s m" boundary
settings in x, y, z directions, respectively. and the box size in z
direction is lz=3.4 A. then in this input file, a restart file was output
through the command of write_restart.
        unfortunately, when i used the output restart file through the
command of read_restart in another input file, i found that the box size in
z direction had been changed to lz=0.000068 A. so i thought the boundary
settings had also been changed to s s s in x, y, z directions, respectively.
        is that a bug in lammps ?

impossible to say. there is not enough information here.
please produce a simple test case that can reproduce this and people
can take a look.

axel.