Software exception monitoring method based on block chain

The invention discloses a software exception monitoring method based on a block chain, a distributed system comprises a plurality of subsystems, the subsystems carry out data interaction with a central database through a switching center, and the method comprises the following steps: collecting network flow data between the interior of the system and an external network and log data in the system as to-be-detected data; after the to-be-detected data of each subsystem is summarized, the system performs analysis and feature extraction on the to-be-detected data based on a deep learning method to obtain possible abnormal data; after possible abnormal data are detected, the data are summarized to an inspection and statistics module to inspect whether the abnormality reported by each subsystem is true or not, and the accuracy of the detection is counted; establishing an alliance chain for system maintenance, and recording statistical abnormal data on the alliance chain; and establishing a consensus mechanism to de