Crunchy Data Warehouse
Power of DuckDB Without Leaving Postgres
In today's data-driven world, the need for versatile database solutions is more critical than ever. Crunchy Data Warehouse was built specifically to address the needs of modern data architectures by integrating DuckDB's query engine directly into Postgres with the fully managed benefits of Crunchy Bridge. This is a cutting-edge solution combining the power and reliability of Postgres with the flexibility and speed of DuckDB.
Postgres + DuckDB
Postgres is renowned for its rock-solid performance, extensive feature set, and proven track record in managing large-scale transactional databases. On the other hand, DuckDB pioneered vectorized execution and as a query engine excels at analytical workloads. By seamlessly integrating DuckDB as a query engine to Crunchy Bridge's Postgres Platform, we provide you with one of the fastest OLAP tools available with the familiarity of a native Postgres experience.
Unified Query Engine
By deeply integrating DuckDB into PostgreSQL, we seamlessly pushdown your complex analytical queries to DuckDB's vectorized engine. You use familiar Postgres tools, while seeing dramatic speed ups in analytics performance.
Simplified Data Management
Easily transfer data between Postgres and cost efficient cloud object store to keep large amounts of historical data for analytics. All without the need for complex ETL processes. Our integration ensures smooth and efficient data flow between the two systems.
Scalability and Flexibility
Continue using Postgres's native Heap tables for responsive transactional performance, while storing your historical analytics data in Parquet format for efficiency and performance or CSV, JSON formats for ubiquitous interoperability.
Lightning fast analytics performance for Postgres
Vectorized Execution
Unlike traditional row-based processing, vectorized execution processes data in chunks or vectors, allowing for significant speed improvements in analytical queries. As a result, you can experience substantial reductions in query execution time, making your data analysis tasks faster and more efficient.
Query Pushdown
Our integration leverages the power of query pushdown to further enhance performance. We recognize the parts of the query plan that can be pushed down into DuckDB and construct the appropriate SQL queries to pass to DuckDB. We use a combination of PostgreSQL hooks to achieve pushdown for complex query operations like filters, aggregates, joins, and more.
Pushdown in parallel
Query pushdown is also executed using parallel execution so complex operations can be split into parts. The parts of the query plan that can be pushed down are converted back to SQL statements and sent to your object store in parallel.
Caching
Files used for queries are automatically cached on high throughput NVMe drives. We also include write-through caching, so as new data is written to the object store, new data is added to the system cache.
Use Cases
Business Intelligence and Analytics
Leverage DuckDB analytical prowess to gain deeper insights from your Postgres transactional data. Easily aggregate and analyze sales and financial data, and perform complex queries without compromising on speed or accuracy. Crunchy Data Warehouse enables your data team to build simple and easy to use tools to deliver actionable insights to decision makers.
Log and Event Analytics
Ingest massive volumes of log and event data, effortlessly perform complex aggregations and transformations, and gain valuable insights into operational metrics and trends. The integrated platform simplifies data workflows, enhances query performance, and provides a unified solution for scalable, high-speed analytics, making it ideal for monitoring system health, detecting anomalies, and driving data-driven decision-making.
Time Series and IoT data Analysis
Store high volume time series data in a cost effective way while creating business value from data analysis across the data set. This integration enables seamless ingestion, storage, and real-time analysis of high-velocity IoT data, providing actionable insights with minimal latency.
Spatial Analytics
Native geospatial support is built into DuckDB and Crunchy Data Warehouse takes full advantage of this. Marry your GeoJSON and GeoParquet data sets with a Postgres and PostGIS front end. The in-memory processing makes spatial data analytics faster and more cost effective than ever before.
Minimize Complexity
Crunchy Data Warehouse eliminates the need for multiple disparate tools, reducing the overhead associated with data movement and format conversion. Users can streamline their workflows and achieve faster insights with less operational complexity.
