Startup Crate.io’s strategy of advancing an open source scale-out SQL database as an alternative to complex NoSQL versions for handling fast-moving machine data appears to be paying dividends with the close of an early funding round and the release of an upgraded version of its platform.
Crate.io this week announced the close of a Series A funding round that garnered $11 million. The round was led by Zetta Venture Partners and Deutsche Invest Equity. Among the other investors is Solomon Hykes, founder of application container pioneer Docker. The funds will be used to accelerate development and adoption of commercial and open source versions of the CrateDB machine data platform, the company said Tuesday (June 19).
The San Francisco-based startup also released the third version of it open source database emphasizing time-series storage and analytics for industrial and other users dealing with large volumes of machine-generated data. The upgrade also targets SQL developers who previously relied on NoSQL approaches to handle machine data applications.
The upgrade also targets users seeking to harness data generated by connected factory equipment along with smart buildings and vehicles. “The capability for real-time processing of machine data [was] a key constraint in many Industry 4.0 endeavors,” noted Torsten Kreindl, managing partner at Deutsche Invest Venture Capital.Industry 4.0 refers to factory automation efforts that incorporate data analytics into manufacturing technologies.
Crate.io said it is addressing the requirements with an upgraded platform that includes faster data ingestion and real-time analytics as well as data visualizations. Along with SQL, it loads JSON and other data points in a variety of structures, including nested objects and arrays.
Meanwhile, data platform administration is based on a cloud-native micro-services approach managed around the Kubernetes cluster orchestrator.
The result is an open source SQL database aimed at real-time processing of Internet of Things and machine data ranging from sensor data and logs to GPS data generated by edge devices. The startup added that the new versionis designed to scale query throughput of complex data structures linearly along with cluster size. That, the startup said, makes capacity both more predictable and affordable.
Other upgrades to version 3.0 include the addition of support for new algorithms such as HyperLogLogthat improve query performance by as much as 100-fold, the company claims. The platform also supports virtual tables and sub-queries.
The upgraded machine data platform is available now.