Crunchy Data Warehouse
Crunchy Data Warehouse is a managed analytics database that extends standard PostgreSQL with native Iceberg table support, vectorized query execution, and parallel processing, delivered as a service on AWS via Crunchy Bridge.
Publisher review
Crunchy Data Warehouse is a managed analytics database that extends standard PostgreSQL with native Iceberg table support, vectorized query execution, and parallel processing, delivered as a service on AWS via Crunchy Bridge. It targets teams already using Postgres who need to run analytical workloads—such as monitoring, IoT, time-series, ad networks, online gaming, observability, user behavior analytics, and geospatial analysis—without migrating to a separate warehouse. The product combines regular Postgres heap tables with fully transactional Iceberg tables backed by S3 object storage, and integrates the DuckDB query engine and other OLAP technologies to deliver over 10x the average performance of PostgreSQL on TPC-H queries.
Users can query raw CSV, JSON, Parquet, Geoparquet, and Iceberg files directly from a data lake, import/export data between Postgres and cloud storage, and write query results back to S3 for advanced pipelines. It supports columnar compression, an adaptive query engine, and automatic Iceberg table maintenance, while maintaining full Postgres compatibility—same SQL, extensions (pg_cron, pg_stat_statements), and tooling. Crunchy Data Warehouse competes with Google BigQuery, Microsoft Azure Synapse, Databricks, Snowflake, and SAP Data Warehouse Cloud, but differentiates by staying on unmodified Postgres rather than forking the codebase, offering a zero-ETL approach that bridges OLTP and OLAP in one environment.
The honest trade-offs include reliance on S3 for Iceberg storage (not a multi-cloud object store abstraction), no explicit mention of real-time streaming ingestion, a managed-only delivery model (no self-hosted option beyond Kubernetes), and a smaller ecosystem of native connectors compared to Snowflake or Databricks. It is available as a managed service on AWS via Crunchy Bridge with pay-as-you-go pricing; on-premises deployments are supported via Crunchy Postgres for Kubernetes. Crunchy Data was founded in 2014 and is headquartered in Charleston, SC.
How it works
-
Iceberg table management
Create, read, update, and manage Iceberg tables backed by S3 object storage with automatic maintenance and version rollback.
-
Vectorized query execution
Uses a vectorized execution engine and parallel processing to achieve millisecond response times on analytical queries.
-
Data lake file querying
Query raw CSV, JSON, Parquet, Geoparquet, and Iceberg files directly from S3 or public URLs without prior loading.
-
Flexible data import/export
Import data from S3 into Iceberg or Postgres tables, and export query results back to S3 for downstream pipelines.
-
Full Postgres compatibility
Supports all Postgres SQL, transactions (DDL and data), schema migration, extensions, and existing tooling without code changes.
-
Zero-ETL analytics
Enables direct querying of data lake files and Iceberg tables, eliminating the need for separate ETL pipelines.
-
On-premises deployment
Available for self-managed clusters via Crunchy Postgres for Kubernetes, supporting on-premises or bare metal environments.
Strengths and trade-offs
Strengths
- Achieves over 10x the average performance of PostgreSQL on TPC-H analytical queries through vectorized execution and parallel processing.
- Maintains full Postgres compatibility—same SQL, extensions, and tooling—so teams can adopt it without retraining or rewriting queries.
- Supports Iceberg tables with automatic maintenance, version rollback, and transactional DDL/data operations on unmodified Postgres.
- Offers flexible data lake integration: query raw CSV/JSON/Parquet/Geoparquet files from S3 and import/export between Postgres and cloud storage.
Trade-offs
- Iceberg storage is tied to S3; no native support for Azure Blob or GCS as a first-class object store backend.
- No built-in real-time streaming ingestion; data must be batched or loaded from files, limiting low-latency use cases.
- Managed service is only available on AWS via Crunchy Bridge; self-managed option requires Kubernetes expertise.
- Smaller native connector ecosystem compared to Snowflake or Databricks, which offer hundreds of pre-built integrations.
Pricing context
Available as a managed service on AWS via Crunchy Bridge with pay-as-you-go pricing; on-premises deployments via Crunchy Postgres for Kubernetes require a separate subscription. Specific tier prices are not publicly listed.
Getting started with Crunchy Data Warehouse
-
Sign up for Crunchy Bridge
Go to the Crunchy Bridge website and create an account. Choose the pay-as-you-go pricing option for the managed service on AWS. Complete the registration process and verify your email address.
-
Connect your AWS account
In the Crunchy Bridge dashboard, link your AWS account by providing IAM credentials or using a cross-account role. This allows Crunchy Data Warehouse to provision S3 buckets and compute resources for your Iceberg tables.
-
Create a data warehouse instance
Select the region and instance size for your warehouse. Configure storage settings, including the S3 bucket for Iceberg data. Review the estimated cost and click Create to deploy the instance.
-
Load data from S3 into Iceberg
Use the SQL interface to import data from CSV or Parquet files in your S3 bucket. Run a CREATE TABLE statement with the USING iceberg clause, then execute COPY or INSERT commands to load the data into Iceberg tables.
-
Run analytical queries
Write standard PostgreSQL SQL queries against your Iceberg tables. Use vectorized execution for fast aggregations and joins. Monitor query performance in the dashboard and adjust instance size if needed.
Frequently Asked Questions
What is Crunchy Data Warehouse?
Crunchy Data Warehouse is a managed analytics database that extends standard PostgreSQL with native Iceberg table support, vectorized query execution, and parallel processing. It is delivered as a service on AWS via Crunchy Bridge, targeting teams who need analytical workloads without migrating from Postgres.
How does Crunchy Data Warehouse improve PostgreSQL performance for analytics?
It integrates the DuckDB query engine and other OLAP technologies with vectorized execution and parallel processing. This combination delivers over 10 times the average performance of PostgreSQL on TPC-H analytical queries, enabling millisecond response times for complex workloads.
Can I query data lake files directly with Crunchy Data Warehouse?
Yes, you can query raw CSV, JSON, Parquet, Geoparquet, and Iceberg files directly from S3 or public URLs without prior loading. This zero-ETL approach allows you to analyze data lake files and import or export data between Postgres and cloud storage seamlessly.
Does Crunchy Data Warehouse support Iceberg tables?
Yes, it supports fully transactional Iceberg tables backed by S3 object storage. Features include automatic maintenance, version rollback, and transactional DDL and data operations, all while maintaining full PostgreSQL compatibility without forking the codebase.
What are the pricing and deployment options for Crunchy Data Warehouse?
It is available as a managed service on AWS via Crunchy Bridge with pay-as-you-go pricing. On-premises deployments are supported through Crunchy Postgres for Kubernetes, requiring a separate subscription. Specific tier prices are not publicly listed.
What are the main trade-offs of using Crunchy Data Warehouse?
Iceberg storage is tied to S3, with no native support for Azure Blob or GCS. It lacks built-in real-time streaming ingestion, and the managed service is only on AWS. The native connector ecosystem is smaller compared to Snowflake or Databricks.
Alternatives
How Crunchy Data Warehouse compares
Direct head-to-head against 3 competitors. Picked by 7wData.
Crunchy Data Warehouse
- Pricing
- Available as a managed service on AWS via Crunchy Bridge with pay-as-you-go pricing; on-premises deployments via Crunchy Postgres for Kubernetes require a separate subscription. Specific tier prices are not publicly listed.
- Target
- Crunchy Data Warehouse is a managed analytics database that extends standard PostgreSQL with native Iceberg table support, vectorized query execution, and parallel processing, delivered as
- Strength
- Achieves over 10x the average performance of PostgreSQL on TPC-H analytical queries through vectorized execution and parallel processing.
- Watch for
- Iceberg storage is tied to S3; no native support for Azure Blob or GCS as a first-class object store backend.
DuckDB
- Pricing
- Free and open source
- Target
- Analytics on local or cloud data lakes
- Deployment
- Embedded or serverless
- Strength
- In-process OLAP engine with zero-config setup
- Watch for
- No built-in Iceberg table management or multi-user concurrency
Dremio
- Pricing
- Free Community Edition; Enterprise starts at $1.50/credit/hr
- Target
- Data lakehouse query engine
- Deployment
- Kubernetes, cloud, on-prem
- Strength
- Native Iceberg support with semantic layer and reflections
- Watch for
- Complex setup and tuning for production workloads
Apache Iceberg (via Trino/Spark)
- Pricing
- Free open source; managed services vary
- Target
- Open table format for data lakes
- Deployment
- Kubernetes, cloud, on-prem
- Strength
- Industry-standard open table format with broad ecosystem
- Watch for
- Requires separate query engine; no built-in Postgres compatibility
User reviews
No user reviews yet. Be the first to write one.
Sources
Reporting on this tool draws on these publicly available sources.