|
Abstract:
Big data has now grown into one of the innovations in the field of data management. Its ability to handle complex data in large sizes in a short period of time becomes an advantages. Especially at this time the development of data in the world of industry and on the internet quite rapidly. The need for big data has encouraged the birth of Hadoop platform which is currently a big data platform that is widely used. As a preliminary study of the implementation of big data, this research took the case of server log analysis. Characteristics of unstructured and rapidly growing log data make it relevant to be a case study of big data implementation. From this research, the Hadoop-based architecture involves HDFS, Yarn and MapReduce, and involves Zeppelin with Pig interpreters to visualize the results of the analysis. Log data is stored in HDFS, and analyzed by MapReduce-based programs. The analysis results are then translated by Pig and then displayed by Zeppelin. The results of the analysis resulting from this study include web access analysis based on IP address, type of browser used, bot's access, daily access, and also weekly access.
|
