Mapreduce frame-based internet of things system

The invention discloses a MapReduce frame-based internet of things system, which comprises a Map job, a MapReduce job, a Reduce work and a distributed data file; one MapReduce job has one or more Reduce jobs; one Reduce work is in charge of reading and processing one data set, the conditions that how many sets are generated and each Map output result is distributed to which set are decided by a final hash function at a Map stage, and each data set is remotely read to a machine where the corresponding Reduce job is at; the Reduce job performs integration processing and analysis on those data and outputs a final data result. The MapReduce frame-based internet of things system can effectively timely process data, can fast accurately analyze a distribution condition of user data and helps a user to fast know data features and make a business/political decision.