8 minute read · April 8, 2024

Dremio’s Commitment to being the Ideal Platform for Apache Iceberg Data Lakehouses

Alex Merced

Alex Merced · Senior Tech Evangelist, Dremio

The data lake and data warehousing space is facing major disruption spearheaded by innovative table formats like Apache Iceberg. Iceberg has now become the cornerstone of modern data architecture. In the Apache Iceberg ecosystem, Dremio has emerged as the frontrunner, championing the use of Apache Iceberg to redefine the potential of data lakes. Dremio has been at the forefront of this journey, hosting dozens of Iceberg talks at the Subsurface conference, authoring the upcoming O’Reilly book, “Apache Iceberg: The Definitive Guide” book, and developing a platform that not only leverages the strengths of Apache Iceberg but also embodies our vision of an open, flexible, and high-performing data ecosystem.

Our journey with Apache Iceberg was not a mere adaptation to industry trends but a deliberate choice aligned with our core belief in open source innovation and our dedication to avoiding the pain in costs and inflexibility of closed data platforms. This commitment has propelled Dremio to become the leading platform for Apache Iceberg, ensuring that organizations can harness the full spectrum of data lakehouse capabilities while maintaining autonomy and cost efficiency.

Historical Alignment with Apache Iceberg

Dremio's alliance with Apache Iceberg dates back to its inception, recognizing early on the transformative potential of this technology. Our initial step towards integration was through the innovative feature of dataset promotion, which epitomizes our approach to enhancing data accessibility and performance. This feature allowed users to elevate file-based datasets, such as a folder of Parquet data, to a structured table abstraction within their data lake, leveraging Dremio's high-performance scan capabilities powered by our Apache Arrow-based query engine.

Integrating Apache Arrow, a project that originated within Dremio, was foundational, transforming how data is processed and queried. Dremio seamlessly creates a layer of Apache Iceberg metadata when a file or directory of CSV/JSON/XLS/Parquet files was promoted to a table within the Dremio platform. This wasn't just about creating a table structure; it was about devising a robust index that facilitates swift and efficient data querying. This strategic move underscored our commitment to Apache Iceberg and demonstrated how deeply integrated it is within Dremio’s architecture.

Feature Examples

In the heart of Dremio’s architecture lies a suite of features that exemplify our commitment to leveraging Apache Iceberg's full potential in every aspect of Dremio’s architecture:

Dataset Promotion: This pivotal feature transforms file-based datasets into structured tables, enabling quick access to raw data. Through Dremio’s dataset promotion, users can effortlessly convert folders of Parquet data into Apache Iceberg tables without rewriting the data or even being aware of the metadata layer, thereby unlocking high-performance querying capabilities. This process is underpinned by Dremio's Apache Arrow-based engine, which accelerates data access enriching the data lakehouse experience with efficiency and speed.

Data Reflections: Dremio’s data reflections eliminate the reliance on materialized views, BI extracts and cubes by creating optimized data structures within the data lake. This feature allows for the automatic substitution of these optimized tables during queries, thus significantly improving query performance. Data reflections are extremely flexible, supporting different sorting, partitioning, and aggregation schemes of the data. All data reflections are now persisted as Apache Iceberg tables in the lake.

Comprehensive Apache Iceberg Support: Dremio doesn’t just integrate with Apache Iceberg; it enhances its utility. With full DDL and DML support, Dremio enables users to seamlessly create, modify, and manage Iceberg tables directly with the simplicity of SQL. This comprehensive support facilitates a direct interaction with Iceberg tables, allowing for sophisticated data operations like ingestion, curation, and analytics within the Dremio ecosystem, thereby elevating the data lakehouse paradigm to new heights of functionality and convenience.

Catalog Versioning and Data Governance

Dremio’s commitment to an open and flexible data platform extends to its catalog versioning and data governance capabilities, highlighted by:

Enhanced Catalog Versioning: Through the integration with Nessie, Dremio Cloud’s integrated Lakehouse catalog introduces robust catalog level versioning for Apache Iceberg tables, enabling features like multi-table transactions, isolated ingestion tasks, and simplified no-copy data replication. This versioning system allows for detailed branch management within the catalog, facilitating complex DataOps workflows and ensuring that data environments are both agile and reliable.

Interoperability and Governance: Dremio Cloud’s lakehouse catalog, powered by Nessie, not only serves its own ecosystem but also ensures compatibility with other engines and tools like Apache Spark and Apache Flink. This interoperability fosters a collaborative data environment where workflows and governance rules can span across various platforms seamlessly, enhancing data governance and enabling a unified approach to lakehouse management.

Simplifying Lakehouse Management

Dremio streamlines the operational aspects of managing a lakehouse by automating Iceberg table optimization and management.

With features like automated compaction and snapshot expiration management, Dremio simplifies the routine tasks of lakehouse upkeep. Users can execute these tasks through straightforward SQL commands like OPTIMIZE and VACUUM, which are integrated within Dremio’s system, making table management both efficient and user-friendly.

These features collectively reinforce Dremio’s position as the ideal platform for Apache Iceberg data lakehouses, providing users with a comprehensive, efficient, and user-friendly environment to manage and analyze their data.


Dremio's unwavering commitment to Apache Iceberg is not merely a strategic choice but a reflection of our vision to create an open, flexible, and high-performing data ecosystem. Our deep integration with Apache Iceberg throughout the entire stack complements Dremio's extensive functionality, empowering users to document, organize, and govern their data across diverse sources, including data lakes,  data warehouses, relational databases and NoSQL tables. This synergy forms the bedrock of our open platform philosophy, facilitating seamless data accessibility and distribution across the organization.

Whether it’s powering BI dashboards through Dremio’s BI integrations or swiftly integrating data into Python notebooks for data science, machine learning, and AI via Apache Arrow Flight, Dremio ensures that data is not just available but optimally utilized.

We invite you to experience the power and versatility of Dremio firsthand. Explore our range of tutorials designed to guide you through the intricacies of the Dremio platform, allowing you to witness its capabilities directly on your laptop. 

From setting up your data lakehouse to executing advanced analytics, these tutorials offer a comprehensive hands-on experience, showcasing how Dremio can transform your data operations and insights. Join us in this journey and see for yourself how Dremio can make your data accessible, manageable, and actionable, wherever and whenever it is needed.

Ready to Get Started?

Bring your users closer to the data with organization-wide self-service analytics and lakehouse flexibility, scalability, and performance at a fraction of the cost. Run Dremio anywhere with self-managed software or Dremio Cloud.