Hadoop Cloud-based Computing Platform Ddos Attack Defense Research

With the development of IT technology, cloud computing has begun to enter into people's vision, it is a revolutionary epoch-making product that integrate the large-scale computing resources through the network and provide reliable services to the user. So far, Google, IBM, Microsoft, Amazon, Alibaba and other world-renowned companies have introduced the corresponding cloud computing platform, and use them as the main development strategy of the future, most of them are using Hadoop as cloud computing base framework.Although the Hadoop framework has become the most popular cloud computing platform in today's society, there is problem that how to deal with the current rampant DDoS attacks, which has become the key factor of self-sustainability of cloud computing development.Firstly, this paper analyzed the principle of DDoS attack and defense, then deeply comprehended the Hadoop cloud computing platform On this base, we advanced a brand new model which based on Hadoop DDoS detection and self- Repair defense with Hadoop own heart monitoring mechanism.Designed a cloud node DDoS detection and health protection threshold function by introducing the hyperbolic tangent function and used this function as the core algorithm. Besides, this paper also designed a DDoS self-repair defense model, then cleaned up the illegal IP address which attacked DDos combined the above model provided real-time data.Finally, through DDoS attacking launching tools attached the Hadoop cloud computing experimental platform , the algorithm was validated. Experimental results showed that the proposed model had good detection results when Hadoop cloud cluster attacked by DDoS It also can well protect cluster node that attacked by DDoS, so that it can effectively prevent the cases of paralysis DDoS attacks, and it did not affect the normal work on Hadoop cluster in the whole process of attack and defense.