Show HN: Postgres and ClickHouse as a unified data stack

ClickHouse pairs Postgres + analytics; HN cheers, compares, grills pricing

TLDR: ClickHouse launched a managed Postgres that streams data into ClickHouse for analytics and lets you query both from Postgres. The community is excited about speed and native joins but is aggressively asking about pricing and how it really differs from Databricks/Snowflake-style offerings.

ClickHouse just dropped a managed Postgres that plugs straight into ClickHouse analytics, promising a “unified data stack” that feels like a power couple. Translation: run your app in Postgres and get fast reports in ClickHouse, with built-in change data capture (CDC — copying changes in real time) and an extension to query ClickHouse from Postgres. They’re flirting with speed too: NVMe (super-fast local storage) to turbo-charge updates and vacuums, plus native joins between your app data and warehouse tables. Private preview is free for now—peek the product and their demo blog.

The crowd? Loud and split. The hype train rolled in with “Looks awesome!” and “more Postgres everywhere” vibes. But the drama hit fast: one commenter immediately grilled pricing, asking if it’s just a markup over Ubicloud. Another poked the bear by comparing this to big-cloud power plays—Databricks and Snowflake—sparking brand war whispers about “what’s actually different” beyond local SSDs. Folks loved the idea of native joins and a Postgres-first workflow, yet threw side-eye at “10x speed” claims and asked for receipts. Memes flew: “Postgres + ClickHouse sitting in a tree,” “NVMe = nitro mode,” and “CDC is gossip for databases.” The mood is: cautiously thirsty for speed, fiercely nosy about price, and ready to debate lock-in vs open-source comfort.

Key Points

  • ClickHouse launched a private-preview managed Postgres service integrated with ClickHouse, built with Ubicloud.
  • The service uses NVMe-backed storage to significantly improve Postgres OLTP performance, with potential up to 10x speedups.
  • Native CDC to ClickHouse is included, currently via ClickPipes/PeerDB using logical replication, with plans for sub-second replication.
  • The pg_clickhouse extension enables unified queries from Postgres into ClickHouse with comprehensive query pushdown.
  • The offering aims to provide a unified data stack: Postgres for transactions and ClickHouse for analytics on an open-source foundation.

Hottest takes

"Looks pretty awesome! Especially the native joins" — scottmas
"similar to how Databricks… or Snowflake… what else is different?" — caffeinated_me
"the more postgres everywhere, the better" — samokhvalov
Made with <3 by @siedrix and @shesho from CDMX. Powered by Forge&Hive.