Analytics Buckets (with Iceberg)

Analytics Buckets provide specialized storage optimized for large-scale analytics using Apache Iceberg open table format. Built on Amazon S3 Tables, Analytics Buckets enable petabyte-scale analytics with automatic compaction and snapshot management.

Key benefits

1. Apache Iceberg format: Open standard with Parquet files for broad tool compatibility.
2. Automatic compaction: S3 Tables merges small files automatically for optimal performance.
3. Built-in time travel: Query historical data using snapshots.
4. Schema evolution: Evolve schema over time without breaking queries.
5. Integrated with ETL: Real-time replication from Postgres via Supabase ETL.
6. Query from Postgres: Use Iceberg Foreign Data Wrapper to join with operational data.

Query tools supported

PyIceberg, Apache Spark, DuckDB, Amazon Athena, and any tool supporting Iceberg REST Catalog API.

Analytics Buckets are valuable for:

• Data tiering: Archive cold data from Postgres to Analytics Buckets.
• Historical analysis: Query complete change history.
• Cost optimization: S3 pricing for large datasets.
• Separation of workloads: Analytics queries don't impact production.
• Bottomless storage: Virtually unlimited capacity.

Security

TLS 1.2+ in transit. SSE-S3 or SSE-KMS encryption at rest.

Analytics Buckets provide the scalable, cost-effective storage layer required for modern analytics workflows.