Data processing method

The invention discloses a data processing method. The data processing method comprises the following steps: obtaining current user behavior data; judging whether the data volume of the obtained user behavior data exceeds the maximum processing capacity of a first data processing application or not, if not, starting the first data processing application to process the user behavior data, and if yes, putting partial data equivalent to the maximum processing capacity in the user behavior data into a normal data queue and putting the rest of data into a retry data queue; and starting the first data processing application to extract the data in the normal data queue and perform data processing, and starting a retry data processing application to extract the data in the retry data queue and perform data processing. According to the data processing method, the situation that subsequent user behavior data cannot be processed in time due to an exception in a processing process is not caused and the user behavior data