基于大数据平台Storm的实时报警系统设计

Design on Real-time Alarm System Based on Large-data Platform Storm

针对系统日志的处理,在传统的大数据框架Hadoop的基础之上,对使用大数据框架Storm进行日志的实时处理进行了研究,设计并实现了一个基于大数据平台Storm的实时日志报警系统。对于符合触发规则的日志,及时地通知管理人员去处理问题。使用3台小型的服务器对该系统进行了仿真,实验表明,系统可以实时地处理系统产生的日志,对于异常的日志可以通过邮件及时地通知负责人。

Aiming at the processing of system logs, this paper studies the realtime using large-data platform Storm based on the traditional large data flame Hadoop, designs and implements a realtime log alarm system based on large data platfonrm Storm. For the logs which are in compliance with the trigger rules, the administrator is notified to deal with the problems in time. The system is simulated using 3 small servers. The experiment results show that the system can process the logs generated by the system in real time, and for the abnormal logs, the person in charge can be notified by mail in time.