الفهرس 
Unifying NoSQL datastores by using an intelligent layer
ContentsOnly 14 pages are availabe for public view
 |  | from | 101 |  |  |
| | | from | 101 | | |
Abstract
Over the year’s cloud computing technology is growing rapidly, the requirements for enormous systems and storage have brought great challenges to relational database management systems (RDBMS) [1]. These important challenges of RDBMS are its inability to deal with Big Data and handle high concurrency applications that demand for high availability, high scalability and high elasticity [2]. Elasticity in specific is contradicted by strict transaction support with ACID properties (atomicity, consistency, isolation, and durability) in traditional RDBMS. These systems are also inflexible to scale to a very big size or with a very high cost. In addition, RDBMS always provide an extremely large feature set which increase cost and complexity however many of the characters are used. Hence, applications that can tolerate reduced or even removed ACID properties are good candidates for using alternate ways of storing data - especially more applicable in cloud environments, where scalability and elasticity requirements are very high.This forced a rising number of companies to improve various types of non-relational database that are in general mentioned to as NoSQL (Not Only Structured Query Language) [3]. A NoSQL technology is a category of database management systems that considered an alternative to relational database; do not store data using traditional fixed table schemas and use different query languages suitable for serving such applications [4]. The main characteristic of NoSQL datastores is that they look for elasticity at the expense of losing SQL integrity capabilities and some processes such as no join; non-relational, easy replication support. An extremely large numbers of different and diverse NoSQL datastores have been developed. The most prominent examples include Apache Cassandra, MongoDB, Voldemort, Apache HBase, SimpleDB, Azure Table Storage, CouchDB, Riak, Redis, and BigTable