Druid Developer Expands Query Options
The latest release of a real-time data analytics platform takes makes use of a new SQL feature in Apache Druid that combines data or rows from multiple tables based on common values.
Imply, the real-time analytics startup founded by the authors of the Apache Druid database, also said it has added a “query laning” feature akin to a carpool lane targeting the most urgent queries, prioritizing resources to handle critical workloads.
Druid, the column-oriented, in-memory OLAP data store, recently added support for SQL JOIN. Imply said this week its version 3.3 release incorporating JOIN would extend Druid’s performance beyond data lake and data warehouse query engines via architectural advantages such as horizontal query distribution and advanced indexing capabilities.
The combination aims to provide the ability to query multiple data sets using standard SQL. doing so without sacrificing performance, Imply said Thursday (April 16).
JOIN support is also touted as promoting self-service analytics by reducing resource requirements like cloud storage volumes and computing overhead. The joining of JOIN with the Druid-based platform also allows “as is” use of multiple data sets, thereby reducing computing and data ingestion costs and ultimately simplifying data pipelines.
For example, existing queries from business intelligence tools such as Looker or Tableau can be used “as is”, without rewriting JOIN queries.
“Our latest release greatly improves the cloud computing economics of real-time intelligence, while maintaining best-in-class performance, so that companies can extend analytics beyond the analyst, to business users, while maintaining fiscal responsibility,” said Fangjin Yang, Imply’s CEO and co-founder.
Meanwhile, the addition of an HOV lane for urgent queries prioritizes access to computing and storage with the goals of improving resource utilization and reducing costs. The framework accelerates time-sensitive queries so they are not stuck in traffic behind, for example, reporting queries.
Query laning also can be used to consolidate mixed workloads onto the same cluster, the startup said.
Founded in 2015 by the creators of Apache Druid, Imply has been ramping up development of its Druid-based analytics engine to deliver real-time queries against huge data lakes. It announced a $30 million funding round in December 2019 led by led by investors Andreessen Horowitz along with Geodesic Capital and Khosla Ventures.
The high-speed query and data ingestion capabilities are aimed at applications like fraud detection, supply chain events and Internet of Things device interactions.