Key-value stores use a simple data model in which data is considered as a set key-value pair; the keys are unique IDs for each piece of data, and they also work as indexes when accessing the data:
- Remote Dictionary Server (Redis): The growing popularity of Redis, an open source key-value NoSQL technology, is a result of Redis's stability, power, and flexibility in executing a wide range of operations and tasks in the enterprise. Redis is used by a diverse set of companies, from startups to the largest technology companies. It is written in C. Redis allows you to conceptualize and approach challenging data analysis and data manipulation problems in a very different manner, as compared to a typical relational data model. In an SQL-based relational database, the developer or database administration create a database schema that organizes the solution domain by normalizing the data into columns, rows, and tables with connecting joins, through foreign-key relationships.
- DynamoDB: DynamoDB is a NoSQL store service provided by Amazon. Dynamo supports a much more flexible data model, especially for key-value stores. The data in Dynamo is stored in tables, each of which has a unique primary ID for accessing. Each table can have a set of attributes that are schema-free, and scalar types and sets are supported. The data in Dynamo can be manipulated by searching, inserting, and deleting, based on the primary keys. In addition, conditional operations, atomic modification, and search by non-key attributes are also supported (yet inefficient), which makes it also closer to a document store. Dynamo provides a fast and scalable architecture, wherein sharding and replication are automatically performed. In addition, Dynamo provides support for both eventual consistency and strong consistency for reads, while strong consistency degrades the performance.