Question 1

How do I sync Postgres to Snowflake?

Accepted Answer

Connect both, pick the tables you want, and set a schedule. Polytomic does the first load in full, then keeps Snowflake up to date from PostgreSQL logical replication. On the Postgres side you need a publication for those tables, a replication slot, and wal_level set to logical. The connecting user needs the REPLICATION attribute, or rds_replication on RDS, plus normal SELECT rights. Polytomic usually creates and drops the slot for you.

Question 2

If a row is deleted in Postgres, does it get deleted in Snowflake?

Accepted Answer

With logical replication, yes. The delete arrives as an event and Snowflake marks the row with a deletion timestamp, so you can still see it and know when it went. If you would rather the row disappear, turn on hard deletes. This is the part worth planning for: the fallback read modes do not carry deletes. A tracking column read and an xmin read both just stop returning the row, so pair those with a full read now and then.

Question 3

Do my Postgres tables need a primary key?

Accepted Answer

Not always, which is more forgiving than the docs suggest. A primary key works, and so does a unique index Polytomic can select. Tables with neither fall back to an xmin read, which behaves like append only data, or to a full read each run. PostgreSQL still needs the right replica identity for updates and deletes to be usable, so the publication and the replica identity have to agree with each other.

Question 4

Will this slow down my production database?

Accepted Answer

Logical replication is designed to be light. Polytomic reads the write ahead log rather than running big scans against your tables, so the ongoing cost is small. The first load does read tables in full, so run it at a quiet time if the tables are large. One thing to watch: a replication slot holds write ahead log while its reader is behind, so a sync left paused for a long time will use disk on the source.

Question 5

Why move Postgres data into Snowflake instead of querying it directly?

Accepted Answer

Because the two jobs pull in opposite directions. Your app database is tuned to serve your product, and large group by queries compete with that. The other half of the problem is that the data you want to join to, like CRM records or billing history, is not in Postgres at all. ETL into a data warehouse moves the heavy queries off production and puts app tables next to everything else, which is where most real questions get answered.

Sync PostgreSQL to Snowflake

Sync PostgreSQL to Snowflake with Polytomic

Deletes come through

Keys set the read mode

Read from a standby

Or sync Snowflake to PostgreSQL

Trusted by businesses worldwide

Step-by-step integration guide

Connect PostgreSQL in Polytomic

Connect Snowflake in Polytomic

Create a sync between PostgreSQL and Snowflake

Map fields and enable the sync

Frequently asked questions

The sync and ETL platform that does it all

Powerful data sync

Advanced analytics

Enterprise security

Easy setup

Ready to connect PostgreSQL with Snowflake?