| With the constant innovation of Internet technology and improvement of basic communication equipment, more and more enterprises and companies are transforming from the traditional offline business to online business. This gives us greater opportunities on one hand, but on the other hand we are also faced with enormous challenges. The explosion of business makes the traditional servers unable to deal with the large-scale data analysis and processing, which leads to low efficiency. At the same time, the increasing business and all kinds of Websites also bring a huge challenge to the network administrator. What’s more, the system security is the primary problems.Cloud computing technology is large-scale, high available, easy-to-get and low-cost. Developed in recent years, cloud computing technology can be used combined with Web log analysis, which can not only deal with massive data processing problems, but also analyze the Website loophole and report to the administrator timely and accurately for quick response in order to get rid of these problems.Through in-depth research of data mining technology and the traces left behind Web attacks, we summed up a general description of Web logs attack. As traditional clustering methods are based on clustering attacks, it cannot work well when an attacker hijacks a large number of computers and uses a similar approach to attack. So we proposed a new cluster algorithm based on clustering attacker purpose and solved the problem that the attacker displayed excessive number.Based on frequently testing the traditional Web log analysis methods and using a timer to analyze the time consumed on every step of Web log mining event, we got the traditional Web log analysis bottleneck, which provided a basis to cloud platform. The results showed that cloud computing didn’t work well when use less logs but it did quite well when there are in large quantity. |
