The rapid development of the fifth generation (5G) promotes a variety of new applications, which will pose a huge challenge to the computing resources of networks. Computing networks is a promising technology, which can provide ubiquitous computing resources for applications in 5G. However, resource optimization in computing networks is still an open problem. In this paper, we propose a novel resource allocation framework for computing networks to investigate the energy consumption minimization problem in terms of delay constraint. To tackle the problem, we propose a dynamic task scheduling and resource allocation algorithm to utilizing the Lyapunov optimization method, which doesn't need to know any prior knowledge of networks. In order to reduce the complexity of solving the problem, we decompose the original problem into several sub-problem to solve. Particulary, the solutions of transmit power and subcarrier assignment are obtained by using the Lagrangian dual decomposition method. The solutions of computation time, postponing time, and CPU-cycle frequency are achieved in the closed form. Simulation results show that the performance of the proposed algorithms and can achieve the tradeoff between the average delay and the average energy consumption.