Big Data in the Cloud: A Survey

Big Data has become a hot topic across several business areas requiring the storage and processing of huge volumes of data. Cloud computing leverages Big Data by providing high storage and processing capabilities and enables corporations to consume resources in a pay-as-you-go model making clouds the optimal environment for storing and processing huge quantities of data. By using virtualized resources, Cloud can scale very easily, be highly available and provide massive storage capacity and processing power. This paper surveys existing databases models to store and process Big Data within a Cloud environment. Particularly, we detail the following traditional NoSQL databases: BigTable, Cassandra, DynamoDB, HBase, Hypertable, and MongoDB. The MapReduce framework and its developments Apache Spark, HaLoop, Twister, and other alternatives such as Apache Giraph, GraphLab, Pregel and MapD - a novel platform that uses GPU processing to accelerate Big Data processing - are also analyzed. Finally, we present two case studies that demonstrate the successful use of Big Data within Cloud environments and the challenges that must be addressed in the future