Nowadays， developments of technology have allowed the generation of huge amounts of streaming data， such as network traffic flows， web click stream， video stream， event stream and semantic concept stream. Therefore， data stream mining has become a hot research topic and its goal is to extract hidden knowledge/patterns from continuous stream data. Clustering， as one of the most important problems in stream mining， has been highly explored recently. However， data stream clustering algorithms differ from traditional static data clustering algorithms in many aspects， and have more constraints such as bounded memory， single-pass， real-time response and concept-drift detection. In this paper， we survey the state-of-the-art data stream clustering algorithms. Firstly， mining constraints are identified. Then a general model for stream clustering is given， and its association with traditional data clustering is described. Finally， some further research issues in this domain are put forward.