The establishment of a real-time monitoring and early warning system (EWS) can effectively reduce the risk of debris flow. Based on the analysis of real-time rainfall data and the automatic identification of rainfall processes, this paper presents a solution and a system for the real-time monitoring and early warning of debris flows (DEWS). Rainfall intensity (RI) and cumulative rainfall (CR) are the main parameters for the debris flow EWS. Thus, determining how to correctly identify a rainfall process is of great significance for improving the accuracy of debris flow EWSs. Based on the store procedure of database and C# development language, combined with the characteristics of rainfall data and the classification criterion of a rainfall process presented by Jan and Lee, the automatic identification of a rainfall process is realized. Developing a method of achieving early warning processes without manual intervention via a completely automatic, real-time and stable operation has represented a difficult problem in the early warning process. The proposed solution introduces “system services” technology to achieve accurate automatic real-time monitoring and early warning for debris flows. This system was applied to the debris flow in Zoumaling gully and successfully predicted the debris flow events on July 8, 2013.