Basho DB Targets IoT, Goes Open Source
The latest installment of a popular NoSQL database geared toward Internet of Things (IoT) applications also has been released to the open source community.
Seattle-based Basho Technologies said Thursday (May 5) the release of its time series Riak database, version 1.3, includes commercial and open source options. The open source version allows free downloads developers can use in production while contributing code and developing applications around Riak TS.
The latest version also is designed to quickly spin up a new instance that can be used to test and build IoT applications, the company said. Version 1.3 also expands support for SQL commands, extends API support and includes support for shell commands as well as multi-cluster replication.
Along with the open-source release, Basho also maintains a laboratory repository on GitHub that includes components such as a front-end application that allows exploration of a Riak cluster. Basho said the open source shift seeks to contribute to industry efforts aimed at addressing distributed infrastructure challenges.
The addition of IoT capabilities to the Basho database underscores the growing amounts of time series data—data marked with a timestamp—from a broad range of IoT sensors and other devices. “The sheer volume of time series data these sources generate requires databases that can efficiently and reliably store and query time series data,” the company noted in a statement.
The latest release of Riak TS aims to allow open source and commercial customers to perform queries and transactions on time series data at scale.
The database is built on Basho’s Riak Core platform, a distributed systems framework based on commodity hardware that emphasizes high availability and “horizontal scalability.”
Among the other version 1.3 upgrades are faster storage and retrieval of time series data achieved by co-locating data in the same physical cluster location based on time range. Meanwhile, the expanded support for SQL commands includes range queries for both unstructured and structured data to make it easier for business users to analyze larger datasets.
Basho also said it is jumping on the Apache Spark bandwagon Apache Spark with a new Spark connector that retrieves data from Riak TS for in-memory distributed processing. The results can then be stored in the time series database as needed. “The ability to persist these results to Riak TS provides flexibility for future data processing or analysis,” Basho asserted.
The upgraded database also includes support for Spark Streaming, DataFrames, the tool for distributed collection of data organized in named columns, as well as Spark SQL.
Meanwhile, the multi-cluster replication feature is designed so time series data can be stored in Riak TS and replicated for analysis and backup clusters, or for multiple datacenters. Replication across datacenters and geographic regions also is designed to comply with regulatory requirements covering stored data.
The database vendor also is betting its customers will need additional horsepower as IoT deployments sweep up more unstructured data. “As enterprises collect IoT time series data from sensors, they will need fast, reliable and scalable read and write performance,” Basho CEO Adam Wray asserted.