apache/incubator-druid
Incubator Druid
Apache Druid is a real-time OLAP database and distributed analytics engine. It functions as a columnar time-series database designed for high-performance analytical queries and the real-time ingestion of streaming and batch datasets.
The system provides a framework for high-concurrency analytics, allowing multiple simultaneous users to execute SQL and native queries across large-scale data. It supports mixed data ingestion, combining real-time streaming and batch loading into a single system for unified analysis.
The platform includes capabilities for distributed cluster management, enabling the monitoring of data sources and system services through a centralized console.
Features
- Time Series Databases - Implements a columnar storage architecture optimized for time-stamped events and rapid aggregation.
- Query Execution - Executes high-performance analytical queries across large datasets using a distributed architecture.
- Analytical Query Engines - Provides a high-performance engine for executing complex analytical queries and reporting on large datasets.
- Data Ingestion - Combines real-time streaming and batch data loading into a single system for unified analysis.
- Inverted Index Engines - Uses bitmaps and inverted indexes to accelerate filtering and aggregation in high-cardinality datasets.
- Distributed SQL Querying - Executes high-performance analytical SQL queries across massive distributed datasets for fast response times.
- Real-time Analytics Platforms - Processes streaming data and executes low-latency analytical queries for immediate operational insights.
- Real-Time Data Streaming - Consumes and indexes events from message buses in real-time for immediate analysis.
- Segmented Storage Architectures - Divides data into time-chunked segments that are replicated across a cluster for parallel processing.
- Persistence & Durability - Provides durable storage for immutable data segments to ensure fault tolerance and system recovery.
- Data Ingestion Tools - Imports both streaming and batch datasets using integrated configuration and monitoring tools.
- Parallel Query Execution - Distributes query fragments across multiple data nodes and aggregates results at a central broker.
- Concurrent Analytical Serving - Serves a large number of simultaneous users performing complex data analysis and reporting.
- Distributed Architectures - Decouples ingestion, storage, and query processing into specialized nodes that scale independently.
- Cluster Management - Offers a centralized console and system tables to monitor data sources and distributed cluster health.
- Relational Databases - Column-oriented distributed store for interactive applications.
