Numerous logs generated during the daily deployment and operation process in cloud computing platforms help system administrators perform anomaly detection. Common anomalies in cloud workflow include pathway anomalies and time delay anomalies. Traditional anomaly detection methods train the learning models corresponding to the two kinds of anomaly detection tasks respectively and ignore the correlation between these two tasks, which leads to the decline of the accuracy of anomaly detection. In order to solve the problems, an anomaly detection method based on multi-task temporal convolutional network was proposed. Firstly, the event sequence and time sequence were generated based on the event templates of log stream. Then, the deep learning model based on the multi-task temporal convolutional network was trained. In the model, the event and the time characteristics were learnt in parallel from the normal system execution processes by sharing the shallow layers of the temporal convolutional network. Finally, the anomalies in the cloud computing workflow were analyzed, and the related anomaly detection logic was designed. Experimental results on the OpenStack dataset demonstrate that, the proposed method improves the anomaly detection accuracy at least by 7.7 percentage points compared to the state-of-art log anomaly detection algorithm DeepLog and the method based on Principal Component Analysis (PCA).

For the overload problem of single controller in Software Defined Network (SDN), a game theory based master controller reselection mechanism-GAME-System Model (GAME-SM) was proposed. Firstly, the problem of switch migration constrained by resource was translated into maximizing revenue problem of zero sum game, and the GAME-SM mechanism was proposed. Secondly, the upper and lower thresholds of the controller load were set to determine the trigger conditions of the game, and the controller whose load reached the upper limit invited the neighboring controllers to participate in the game as game players. Finally, the game strategy was designed based on the zero sum game to maximize the revenue of each participant, and the master controller was reselected by the repeated game with the change of utility degree, and the load balance of the whole system was realized eventually. The simulation results show that the proposed mechanism can significantly improve the controller load balance, and the controller response time is reduced by 50% compared with Distributed-CoNTroLler (D-CNTL).

Most of the controller deployment programs in Software Defined Network (SDN) are focused on the propagation delay under normal network state, ignoring link fault state on the delay. To solve these problems, a controller deployment scheme based on delay optimization was proposed. Firstly, based on the worst delay minimization problem under normal network states and multiple single-link fault states, the network state delay was used as the new delay optimization goal to establish a controller deployment model. Secondly, two heuristic deployment algorithms were proposed to solve the above model:Controller Placement Algorithm based on Greedy Algorithm (GA-CPA) and Controller Placement Algorithm based on Particle Swarm Optimization (PSO-CPA). Finally, in order to measure the performance of the solutions, some real network topologies and data were chosen to verify the validity. The simulation results show that GA-CPA and PSO-CPA algorithms can reduce the network state delay in different degrees, thus ensuring that the worst delay in most network states is maintained at a lower range.