Distributed Denial-of-Service (DDoS), is one of the most serious threats to the server and network. Having a deeper understanding of DDoS patterns is crucial to build defense and filter systems to guard the network. This paper, using the CICDDoS2019 dataset, is trying to perform data analysis over common DDoS of various kinds. An exploratory data analysis is applied to the dataset to reveal the general characteristics of DDoS attacks and select features; and several models including SVM, random forest, and multiple layer perceptron are compared for their performance on differentiating normal network traffic and DDoS packets. Random forest is recognized as the best model against DDoS, with 99.99% accuracy using selected features.