Data movement¶
Snowflake Postgres offers two complementary ways to move data between your Postgres instance and Snowflake. Both are built into the platform — no external CDC service, no connectors, no separate infrastructure to run.
- Data mirroring for continuous, whole-table replication into Snowflake with schema evolution and a queryable change feed.
- pg_lake for selective, on-demand data movement between Postgres and Iceberg tables, stages, or customer-managed cloud storage.
Compare your options¶
| Data mirroring | pg_lake | |
|---|---|---|
| What moves | Whole tables, continuously | Specific data you select |
| Direction | Postgres → Snowflake (currently) | Bidirectional |
| Latency | ~30 seconds via $live views; configurable syncs | On-demand (you control when data moves) |
| Use case | Analytics replica of operational data | Selective sync, file exchange, Iceberg tables |
You can use both in the same account — they solve different problems.
Data mirroring¶
Data mirroring continuously replicates tables from Snowflake Postgres into a Snowflake analytics database. Every insert, update, delete, and schema change on the source appears on the target automatically, without ETL pipelines, batch jobs, or third-party connectors.
Use it when you want the analytics side to look like a live, always-fresh copy of the OLTP side.
Learn more: Data mirroring
pg_
pg_lake is a Postgres extension that lets you selectively move data between Postgres and object-storage formats like Parquet and Iceberg. With pg_lake, you can create Iceberg tables directly in Postgres, sync individual datasets to Snowflake, or exchange files bidirectionally through stages or customer-managed cloud storage.
Use it when you need direct control over what moves and when, or when you need file-based exchange or bidirectional data movement.
Learn more: pg_lake