Big Data Storage - Big Data: Related Technologies, Challenges and Future Prospects - page 38

to store the large amount of small-sized photos. Taobao also developed TFS and
FastDFS. In conclusion, distributed file systems have been relatively mature after
years of development and business operation. Therefore, we will focus on the other
two levels in the rest of this section.

4.3.1 Database Technology

The database technology has been evolving for more than 30 years. Various database
systems are developed to handle datasets at different scales and support various
applications. It is apparent that traditional relational databases cannot meet the
challenges on categories and scales brought about by big data. NoSQL databases
(i.e., non traditional relational databases) are becoming more popular for big data
storage. NoSQL databases feature flexible modes, support for simple and easy
copy, simple API, eventual consistency, and support of large volume data. NoSQL
databases are becoming the core technology for of big data. We will examine
the following three main NoSQL databases in this section: Key-value databases,
column-oriented databases, and document-oriented databases, each based on certain
data models.

4.3.1.1 Key-Value Databases

Key-value Databases are constituted by a simple data model and data is stored
corresponding to key-values. Every key is unique and customers may input queried
values according to the keys. Such databases feature a simple structure and the
modern key-value databases are characterized with high expandability and smaller
query response time higher than those of relational databases. Over the past few
years, many key-value databases have appeared as motivated by Amazon's Dynamo
system [ 9 ]. We will next introduce Dynamo and several other representative key-
value databases.

Dynamo

Dynamo is a highly available and expandable distributed key-value data storage
system. It is used to manage store status of some core services in the Amazon
e-Commerce Platform. Amazon e-Commerce Platform provides multiple services
and data storage that can be realized with key access. The public mode of relational
databases may generate invalid data and limit data scale and availability. Dynamo
can meet requirements of such applications with a simple key-object interface.
The Dynamo interface is constituted by simple reading and writing of data items.
Dynamo achieves elasticity and availability through the data partition, data copy,
and object edition mechanisms.