This paper proposes a method to solve the joint optimization of test node selection and fiber thread connection. The problem is modeled as an integer linear programming problem with two classes of decision variables. The first class of decision variables are composed of the number of parallel fiber threads throughout each potential testing route, the other class of decision variables are imposed just to denote whether a node is decided as a location to place testing equipment. The objective of the joint optimization is to minimize the number of nodes where fiber test equipment based on the optical time domain reflectometer are planned to locate. The objective function is just the sum of the second class of decision variables. Two types of linear constraints are needed. The first type of constraints describes the mutual relation between the two classes of variables, and the other type of constraints are formed to depict the impact of the number of fiber threads mounted within each fiber link on the first class of decision variables. Multiple examples are given to exhibit the advantage of joint optimization over heuristic method.