Atomistic simulation methods are appropriate tools for investigating the dynamics of dislocations and their interactions with obstacles in metallic materials. In particular, molecular dynamics (MD) simulations have been widely employed on these two topics in the past several decades. However, even for the same type of simulation, the results can vary. While some of the quantitative differences may be due to the choices of interatomic potential and simulation cell size, they could similarly be attributed to choice of model settings, which have also differed substantially to date. In this paper, we carry out systematic MD simulations to study the effects of a few key model settings on the dynamics of an edge dislocation and its interaction with a void in copper. For a fixed interatomic potential, three modeling parameters, including applied loading mode, boundary conditions, and thermostat, are considered and their influences on the stress–strain response, the dislocation velocity, and the critical stress for a dislocation to bypass a void are compared. For a few select cases, we further examine the influence of temperature, strain rate, and simulation cell size. The results show that (i) compared with flexible boundary conditions, rigid boundary conditions result in greater stress oscillations in simulation cells of certain sizes; (ii) compared with the cases of no thermostat and a full thermostat, a partial thermostat provides better temperature control and lower friction on the dislocation core, respectively; and (iii) for dislocation–void interactions, the critical dislocation bypassing stress in shear loading can be appropriately determined with either a constant applied strain rate or a constant applied stress although the strain rate cannot be controlled in the latter. This analysis reveals that these three settings greatly influence the accuracy and interpretation of the results for the same type of simulation.