Based on the general requirement of the existing network centralized alarm monitoring and fault handling

the technology architecture of intelligent monitoring system based on Spark/Hadoop was proposed which can dig and analyze massive alarm data. A distance-based alarm big data correlation analysis algorithm was mainly studied. Through analyzing the actual data

this method can automatically and intelligently correlate the alarms and aggregate the rules of alarm correlation so as to enhance the validity of the rules and improve the alarm compression ratio and the trouble-shooting ratio.