August 9, 2026
WAL Street meets comment chaos
How We Pushed CDC into Postgres
Snowflake says database syncing is finally ‘push-button’—but commenters say, ‘Been there’
TLDR: Snowflake says it made live database copying into its analytics system far simpler and more reliable, turning a painful setup into a push-button feature. Commenters were split between impressed users, rivals saying older tools already did this, and skeptics questioning how “open” Snowflake’s open-source pieces really are.
Snowflake came in swinging with a big promise: press a button and your Postgres database magically starts showing up in Snowflake, neatly and reliably, without the usual nightmare of brittle tools and babysitting. In plain English, they’re trying to make the messy job of copying live app data into an analytics system feel less like a Rube Goldberg machine and more like clockwork. That part got at least one full-on fan response, with one commenter calling Snowflake “a really amazing product” and saying it’s been “a delight” to use.
But the comment section quickly turned into the classic tech food fight: is this actually new, or just newly packaged? One veteran immediately hit the brakes, saying this was basically Vertica’s old “party trick,” while another name-dropped Oracle GoldenGate doing similar magic back in 2014. Translation for non-database mortals: several readers were basically yelling, “Congrats on inventing something other companies already sold us a decade ago.” Ouch.
Then came the rival fandom energy. A ClickHouse supporter jumped in to say PeerDB already “nailed this,” bragging that they used it on giant databases and “essentially never thought about it” — which, in infrastructure land, is about as romantic as it gets. And for the open-source crowd, there was side-eye over Snowflake’s pg_lake story: yes, it’s open source, but one commenter pointed out the juicy replication piece appears closed, sparking the always-reliable ‘open… but how open?’ debate. So yes, Snowflake brought the big launch vibes — but the real show was the comments, where praise, nostalgia, vendor rivalry, and open-source suspicion all started throwing elbows.
Key Points
- •Snowflake introduced a new Snowflake Postgres feature called data mirroring, available in public preview.
- •The feature is designed to replicate Postgres data into Snowflake with low lag, low cost, transactional consistency, and no extra infrastructure.
- •The article argues that traditional Postgres CDC pipelines are fragile because clients must manage schema changes, snapshots, failures, and transaction boundaries.
- •Snowflake’s approach pushes changes directly from Postgres into Apache Iceberg tables stored as compressed Parquet files.
- •The article presents object storage destinations such as Amazon S3 as reliable CDC targets and contrasts this push model with Postgres logical decoding’s client-heavy pull model.