This paper gives overview of two modern approaches to enterprise resources scheduling problem: continuous time problem setting that leads to a combinatorial set of classical linear programming models and discrete time approach that gives an integer programming model. We follow the main steps of problem formulating and find decision points where the researcher chooses which model and computational method to apply. The main objective of this research is to compare both approaches from the computational complexity point of view. We consider several special cases of mass and individual manufacturing and define which approach is more applicable for each of them. In practice this advice can be used in particular cases to reduce hardware requirements to the enterprise resource planning (ERP) system and in certain cases can help improve its performance avoiding implementing big data techniques.