Install gravitino with Homebrew

Project links

• Homepage
• Repository
• Upstream docs
• Package manager page

Commands and aliases

• gravitino-iceberg-rest-server.sh
• gravitino.sh

Project history

Gravitino was created to address metadata fragmentation across heterogeneous data and AI systems. The Apache project blog describes it as a unified metadata layer, or metalake, spanning systems such as Apache Iceberg, Apache Hive, Apache Kafka, MySQL, PostgreSQL, HDFS, S3, and model metadata.

The project entered the Apache Incubator in June 2024 and published its first ASF release, 0.6.0-incubating, on September 9, 2024. That release emphasized enterprise catalog concerns such as RBAC, Iceberg REST catalog modularization, tagging, and connector work.

Apache announced Gravitino as a Top-Level Project on June 3, 2025, citing a growing community and production adoption. The 1.0.0 release on September 24, 2025 framed the project as moving from metadata management toward metadata-driven action systems, including statistics, policies, jobs, MCP integration, lineage, and governance.

Adoption history

The official graduation announcement names production or deployment interest from companies including Pinterest and Uber, and describes adoption by organizations with serious data challenges. That adoption path is tied to open table formats, multi-cloud metadata, and the need to govern assets across engines.

Gravitino's Homebrew packaging makes it easy for data platform engineers to start a local server, test the Iceberg REST server scripts, or evaluate metadata APIs without first building the Java project from source.

How it is used

The package exposes gravitino.sh and gravitino-iceberg-rest-server.sh, reflecting two common local workflows: run the Gravitino server or run the Iceberg REST catalog service. Official installation documentation points users at conf/gravitino.conf for server settings and conf/gravitino-env.sh for environment settings.

Operationally, Gravitino is not just a command. It is a service surface with catalogs, schemas, tables, filesets, topics, models, connectors, authorization, and REST APIs that data platform teams integrate with query engines and governance systems.

Why package nerds care

For package nerds, Gravitino is a reminder that Homebrew can carry data platform services, not just desktop tools and Unix utilities. The formula's value is in reproducible local evaluation of a JVM server and related scripts.

Its Apache lifecycle also matters: incubating release names, Top-Level Project graduation, Java runtime requirements, and generated distribution layouts all affect how downstream package managers name, test, and upgrade the package.

Timeline

• 2024-06: Gravitino enters the Apache Incubator.
• 2024-07: The Apache blog introduces Gravitino as an open-source data and AI multi-cloud solution.
• 2024-09: Apache Gravitino 0.6.0-incubating is announced as the first ASF release.
• 2025-05: Apache Gravitino 0.9.0 focuses on AI, data governance, and security features.
• 2025-06: Apache announces Gravitino as a Top-Level Project.
• 2025-09: Apache Gravitino 1.0.0 is announced with metadata-driven action systems and MCP server work.

Related projects

• Apache Iceberg: Gravitino provides and packages Iceberg REST catalog server functionality.
• Apache Hive, Kafka, Spark, Flink, Hudi, Paimon, Ranger, MySQL, and PostgreSQL appear in official Gravitino materials as systems it integrates with or governs.
• Datastrato is part of the project's origin story and ecosystem around the metadata lake idea.

Sources

• https://github.com/apache/gravitino
• https://gravitino.apache.org/blog/gravitino-0-6-0-release-notes/
• https://gravitino.apache.org/blog/gravitino-0-9-0-release-notes/
• https://gravitino.apache.org/blog/gravitino-1-0-0-release-notes/
• https://gravitino.apache.org/blog/gravitino-open-source/
• https://gravitino.apache.org/blog/gravitino-top-level-project/
• https://gravitino.apache.org/docs/0.9.1/how-to-install/