Apache Druid

Apache Druid is a high-performance, real-time analytics database designed to deliver sub-second query responses on streaming and batch data at massive scale and under sustained concurrent load. It excels at powering interactive analytics applications that require millisecond-level query latency on datasets ranging from billions to trillions of rows, with high-cardinality and high-dimensional data that would overwhelm traditional OLAP systems. Druid's architecture utilizes a scatter-gather query engine with data preloaded into memory or local disk, enabling it to support hundreds to hundreds of thousands of queries per second at consistent performance while using less infrastructure than competing solutions. The system provides native integration with Apache Kafka and Amazon Kinesis for real-time streaming ingestion, supporting query-on-arrival at millions of events per second with guaranteed consistency, while also supporting batch loading from HDFS, Amazon S3, Google Cloud Storage, and other data lakes. Key features include automatic data summarization through roll-up, approximate algorithms for fast count-distinct and quantile calculations (HyperLogLog, DataSketches), multi-tenant query isolation with priority-based scheduling, and column-oriented storage with bitmap indexes for rapid filtering. Druid is trusted by major organizations for real-time analytics use cases including ad tech dashboards, network telemetry monitoring, digital marketing analytics, and business intelligence. Companies like Netflix, Airbnb, Walmart, and Slack rely on it to power user-facing analytics products that serve millions of concurrent users. As an Apache Software Foundation top-level project, Druid is open-source under the Apache License 2.0 and benefits from an active community of hundreds of contributors.