Basho introduced Riak Enterprise 2.0. The company is the creator of Riak, a distributed NoSQL database.
The two big additions in version 2.0 are enhanced search functionality and advanced data types. The update also comes with some security enhancements and simplified configuration management. The enhancements position Riak 2.0 as a platform for a wider variety of applications.
The previous search functionality was built in-house, but customers demanded more full-featured capabilities, according to Peter Coppola, vice president of product at Basho.
Apache Solr has been leveraged for search in 2.0 giving better performance. “With this powerful search capability, we can now support different apps within the organization,” said Coppola. “It moves us from point solution to platform.”
The 2.0 release also includes support for new distributed data types, including sets, flags, register and maps.
Here are some use cases:
- Sets: An e-commerce company may use this to represent items in the shopping cart, or the list of connections on a social network.
- Flags: Detecting when something is retweeted, identifying whether someone is a premium user on LinkedIn.
- Registers: A list that changes value over time, such as what’s trending, top 10 or favorites.
- Maps: A combination of the above data types to create a data structure for a user profile.
The addition of these data types means users don’t have to write the code for them and brings better conflict resolution for a more consistent database. “We’ve defined a conflict resolution methodology,” said Coppola. “That conflict resolution may differ from data types.”
Other upgrades include:
- Security: Riak now enables authorization and authentication to manage users and groups. Security measures can now be applied within Riak itself, including data access and functional permissions.
- Simplified configuration management: To add further operational simplicity, configuration information has been further consolidated and stored in an easy-to-parse, transparent format.
- Efficient bucket management: with bucket types, users can create and administer bucket properties and apply them to a collection of buckets, improving overall efficiency.
- Tiered storage: Riak allows LevelDB users to split data files across two mount points based on access patterns to optimize for low latency of the most frequently accessed data.