Abstract
The influence of the two fast evolving paradigms, Big Data and Cloud Computing, is driving a revolution in different fields of computing. The field of databases is not an exception and has been influenced profoundly by these two forces. Cloud computing is adding to the drive towards making the database available as a service on the cloud. It is shifting the traditional ways in which data is stored, accessed, and manipulated with the appearance of the NoSQL concept and domain specific databases, consequential in moving computing closer to data. This chapter provides a general idea of the changes brought about by the upcoming paradigms in database storage, management, and access, and also provides a brief account of the recent research in the field.
TopIntroduction
With cloud computing taking center stage more and more businesses are making an allowance for making the switch from the physical to the virtual. Increased access to information and empowerment of users is the key to the qualitative benefits provided by the Cloud. With the availability of data in the cloud, as the Oracle white paper by Greenwald (2012) affirms, users would be able to produce more value from their data based on increased flexible access to that data, letting the data to be collectively processed with their domain expertise to produce real and important business benefits.
According to the McKinsey Global Institute's 2011 report on Big Data authored by Manyika, Chui, Brown, Bughin, Dobbs, Roxburgh and Byers (2011), almost all sectors in the United States of America have, on average, hundreds of terabytes of data accumulated per company. Several of these companies have even by now exceeded the 1 petabyte mark. And as the tools and technologies of data storage and management evolve, the volume is only going to amplify in multiples.
Mayer-Schönberger, Cukier (2013) in their recent publication, ‘A Revolution that will change the way we Live, Work and Think- Big data’, have very lucidly described the ongoing transformations in the digital data sector. Digital data they say, doubles a little more than every three years. In the context of Big Data they have introduced a new term, Datafication, that refers to taking information about all things and transforming into a data format to make it quantified so as to use the information in new ways, as in predictive analysis, so as to unlock the implicit latent value of information. These future directions about the use of data point towards a situation where the availability of data may always be an asset and so may its relevance.
As the data flows in from all directions, decision making will further be influenced by the quantitative and diversity dimensions of data. Data therefore will have to be available anytime-anywhere and every-time everywhere.
With everything hosted in the cloud nowadays, hosting of databases on the cloud is but a natural option. With the business switching to cloud and the increase in demand, volume, and need for analysis of data, effective management of data in the cloud environment is imperative. Therefore there has been lots of interest in research in the database management sphere since the inception of cloud computing, to study its integration with the environment. The need for scalable database i.e. database capable of expanding to accommodate growth, has increased with the growing data in the web world. Web applications that need to store and retrieve data for very large numbers of users have been a major driver of cloud-based databases. The needs of these applications are different from those of traditional database applications, since they value availability and scalability over consistency. With increasing volume and complexity of data, evolving technologies, and changing needs of the consumers, study of cloud based databases or cloud databases, is catching increasing attention. This chapter provides a description of the fundamentals of cloud databases and is organized in nine sections. Section 1 gives an introduction to the topic, section 2 describes the basics of cloud databases and section 3 deals with their components and architecture. In section 4, the Data Models for Cloud Databases are described. Sections 5 through 7 provide a broad description of the Data Models. Section 8 deals with the recent research in the cloud database domain and Section 9 concludes the chapter.
TopCloud Database
There is no clear definition for the term ‘Cloud database’, though it seems so easy to understand if one knows about the ‘cloud’ environment and a ‘database’. Publications on cloud computing and databases also indicate that the definition of what a cloud database actually is, is somewhat unclear. More than what it is, what it is not is clearer. A cloud database is not merely taking a traditional RDBMS and running an instance of it on a cloud platform.
Some researchers have made an attempt to define cloud databases in their own contexts. Some of these descriptions are,
Key Terms in this Chapter
Document Stores: Collections of documents of any length that allow retrieval of data based on the document content.
SQL Databases: Databases that use the standard query language (SQL) for the users’ interaction with the database. RDBMSs have SQL as their integral component.
Database Scalability: The ability of a database to meet the escalating demands generated due to increase in volume of data.
Wide Column Stores: Column-oriented databases; a column-oriented database stores its content by column rather than by row. Column oriented databases are a hybrid of classic relational databases and the column oriented technology.
NoSQL Databases: Databases which use Not Only SQL, but any other language for accessing the data. The data therefore may not be structured, as it is in RDBMSs.
Key-Value Stores: NoSQL databases where a key points to a value that is usually a random string.
Data-as-a-Service (DaaS): A cloud based service that provides data to the consumer on demand.
BASE Properties: The relatively lenient properties that a database may possess. These include the properties of Basic Availability , Soft state and Eventual consistency .
Unstructured Data: Data that either does not have a pre-defined data model or is not organized in a pre-defined manner.
Graph Databases: Databases that use graph structures with nodes, edges and characteristics to depict and store information.
Cloud Database: An optimized database storage, management and retrieval service delivered on demand to the users, through the Internet from a cloud database provider’s server.
Cloud Storage: The storage of data in the virtual storage of the cloud.
Hybrid Databases: Hybrids of SQL-NoSQL database solutions that combine the advantage of being compatible with many SQL applications and of providing the scalability of NoSQL.
Relational Database Management System (RDBMS): RDBMS is based on the relational model prescribed by E. F. Codd.
Atomicity, Consistency, Isolation, and Durability (ACID) Properties: These are a set of features that ensure that database transactions are carried out reliably. Atomicity is the ability of the database to guarantee that either all of the tasks of a transaction are performed or none of them are. Consistency is the property that ensures that the database remains in a consistent state before the start of the transaction and after the transaction is over. Isolation refers to the requirement that other operations cannot access or see the data in an intermediate state during a transaction. Durability refers to the guarantee that once the user has been notified of success, the transaction will persist, and not be undone.
Data Management: This comprises all the regulations associated with managing data as a valuable resource.
Database-as-a-Service: A cloud based service that provides database functionalities to the consumer on demand.