NoSQL databases - no panacea for Big Data processing

Big Data, big number of users, and cloud computing are driving the adoption of new database architectures, particularly NoSQL databases. Both research and practice indicate that traditional universal DBMS architecture hardly satisfies new trends in data processing in such environment. NoSQL databases enable better application development productivity through a more flexible data model, greater ability to scale dynamically to support more users and data, an ability to develop highly responsive applications and more complex processing of data. On the other hand, NoSQL databases support solving data problems only partially. We will describe their basic features like horizontal scalability and concurrency model, which offer mostly weaker tools for querying and transactions processing than relational SQL-like database systems do. We will also present some data models and querying capabilities of NoSQL databases in more detail as well as an overview of some their representatives. The NoSQL system properties mentioned imply that most of them are unsuitable, e.g., for the DW and BI querying or, in general, for the enterprise data processing. Consequently, new database architectures and various hybrid solutions are developed. We will point out on these actual problems and present some of the current approaches in detail.