Use Cases & Architecture Patterns
pg_mooncake enables four high-value analytical patterns directly inside PostgreSQL without introducing Kafka, Debezium, or an external data warehouse.
Write to your primary OLTP rowstore table and route slow analytical aggregations to a sub-second USING mooncake columnstore mirror.
Mirror a time-partitioned PostgreSQL table into object storage, drop old partitions from the rowstore to save disk space, and retain full queryable history in the columnstore.
Automatically export PostgreSQL tables as Apache Iceberg v2 tables on S3 or Cloudflare R2, and query external Iceberg/Delta Lake datasets inside Postgres.
Ingest high-volume application logs and traces into PostgreSQL and run fast vectorized analytical queries over object-storage-backed Parquet files.
Speed up slow analytical queries on PostgreSQL tables by routing reads to a real-time USING mooncake columnstore mirror while keeping your OLTP write path unchanged.
Archive time-partitioned PostgreSQL tables into Parquet and Apache Iceberg storage with pg_mooncake, dropping cold partitions from the rowstore while retaining full history in the columnstore.
Build a bidirectional Apache Iceberg v2 lakehouse in PostgreSQL: export operational Postgres tables to S3/R2 without Kafka or Debezium, and query external lakehouse tables in SQL.
Store and analyze high-volume application logs and OpenTelemetry traces in PostgreSQL with pg_mooncake and object-storage-backed Parquet columnstore mirrors.