NoSQL database systems are today an effective solution to manage large data sets distributed over many servers. A primary driver of interest in NoSQL systems is their support for next-generation web applications, for which relational DBMSs are not well suited. These are OLTP applications for which (i) data have a structure that does not fit well in the rigid structure of relational tables, (ii) access to data is based on simplereadwrite
operations, (iii) scalability and performance are important quality requirements,
and (iv) a certain level of consistency is also desirable [7, 20]. NoSQL technology is characterized by a high heterogeneity [7, 21], which is problematic to application developers. Currently, database design for NoSQL systems is usually based on best practices and guidelines [12], which are specifically related to the selected system [19, 10, 17], with no systematic methodology. Several authors have observed that the development of high-level methodologies and tools supporting NoSQL database design are needed [2, 3, 13].In this paper we aim at filling this gap, by presenting a design methodology for NoSQL databases that has initial activities that are independent of the specific target system. The approach is based on NoAM (NoSQL AbstractModel), a novel abstract data model for NoSQL databases, which exploits the observation that the various NoSQL systems share similar modeling features. Given the application data and the desired data access patterns, the methodology we propose uses NoAM to specify an intermediate, system-independent data representation.The implementation in target NoSQL systems is then a final step, with a translation that takes into account their peculiarities

by :Francesca Bugiotti1∗, Luca Cabibbo2, Paolo Atzeni2, and Riccardo Torlone2

File information :English langauge ,8 page,Size :99.8 k

Download link :click

کاری از :Francesca Bugiotti1∗, Luca Cabibbo2, Paolo Atzeni2, and Riccardo Torlone2

اطلاعات فایل :زبان انگلیسی /8صفحه /حجم :99.8k

لینک دانلود :روی همین لینک کلیک کنید