Big Data The NoSQL Column Store Implementation Uses Apache Accumulo to Build a Big Data Infrastructure Model
The NoSQL Column Store Implementation Uses Apache Accumulo to Build a Big Data Infrastructure Model
AbstractAs the development of the digital world and technology, making data growth becomes very fast. This fast data growth gave rise to a new phenomenon, Big Data. However, the database that has been used, namely RDBMS has limitations in managing Big Data. Therefore, in this final project a Big Data infrastructure model will be built by implementing NoSQL using a single node and multi node cluster Apache Accumulo with seven nodes. The infrastructure that has been built will be tested for performance based on the results of runtime, throughput and latency using YCSB. The test is done based on variations in data size (500MB, 1GB, 1.5GB and 2GB), variations in the number of nodes (1, 4, 5, 6 and 7 nodes) used and variations in testing time (morning, afternoon and night). Testing uses YCSB core workload, namely workload A, B, C, D, E, and F which consists of two processes, load and run. From the results of testing and analysis of data readings obtained, runtime is influenced by the throughput and latency generated. The optimal infrastructure model is a multi node cluster using 4 nodes and the optimal time is at night.