Apache Gravitino: A Multi-regional, Geo-distributed Meta Datalake

on YouTube ↗31m

" summary

Apache Gravitino presents a geo-distributed metadata lake that discovers schemas automatically, propagates schema changes back to sources, and runs cross-source SQL joins (demoed joining Hive and Postgres) without moving terabytes. Built connectors include Hive, Iceberg, Hoodie, JDBC-backed databases, Kafka streaming and a custom Trino connector; it uses Apache Ranger for access control, emits OpenLineage lineage, and exposes REST / CLI / Java / Python APIs. Recent v0.9 adds AI model and file-catalog support, a virtual file system, Helm/Kubernetes charts, and ~197 contributors with 1.0 forthcoming.

AI-written summary · about

" chapters (14)

1. 0:00Introduction
2. 1:04What is Apache Gravitino
3. 3:57Key features
4. 7:46Demo setup and API
5. 10:16UI and CLI interaction
6. 13:10Metadata entities management
7. 16:27Catalog types and sources
8. 18:24Programming APIs and hooks
9. 19:11Multi-cloud and multi-region
10. 20:12Cross-source queries demo
11. 22:45Resources and documentation for Gravitino
12. 23:49Use cases and benefits
13. 24:30Project status
14. 25:13Q&A on failover

" related talks

• 33mStreamlining Metadata Management with PostgreSQL and Apache GravitinoLisa Cao · Postgres Conference Orlando 2025
• 37mBreakdown New Data Silos in Generative AIJunping Du · AI.dev: Open Source GenAI & ML Summit North America 2023
• 34mApache Iceberg.Dipankar Mazumdar · Scale By the Bay 2023
• 50mDesigning a scalable data platform based on Apache Spark/IcebergStephen Said, Norman Weisenburger · Voxxed Days Luxembourg 2024
• 34mCloud Agnostic Data Platform with DBT, Trino, Iceberg and MinIOLeuci, Calabrese · GeeCON 2023
• 37mApache Iceberg: An Architectural Look Under the CoversAlex Merced · Øredev 2025
• 31mBuilding large scale, transactional data lakes with Apache HudiNishith Agarwal · Big Data & AI Toronto 2020
• 38mHerding Cat(alogue)s: The Apache Iceberg™ Catalog LandscapeDanica Fine · Berlin Buzzwords 2026