We wrote a guide + demo showing a code‑first path from Postgres to ClickHouse using Debezium and Redpanda, with MooseStack as the dx/glue layer.
This post targets teams who want code ownership: typed streams, stateless streaming transforms, and OLAP tables defined in code, with a local dev server to spin up ClickHouse + Redpanda + Debezium + Postgres in one command.
Blog is linked in the repo.
Right now the demo solely focuses on the local dev experience, looking for community input regarding best practices for running Debezium in production.
oatsandsugar•3mo ago
I think my favorite part was the ability to use the same Drizzle TS data models created for Postgres for creating tables in ClickHouse
Bluegoop•3mo ago
Super cool everything is better in code! :-)
tacone•3mo ago
Very nice. Is this only possible with postgres or might be possible with sqlite or some sqlite variant?
oatsandsugar•3mo ago
Co-author here, we used Debezium here because it supports many different databases. Unfortunately, no sqlite support—my understanding is as an embedded db it lacks some of the prerequisites that Debezium relies on.
Have you run CDC from sqlite? would love to hear how you did it and to try build a demo with MooseStack
tacone•3mo ago
Hello, thank you for your answer. No I never ran CDC from sqlite, I am actually just investing sqlite, thus my question.
okane•3mo ago
This post targets teams who want code ownership: typed streams, stateless streaming transforms, and OLAP tables defined in code, with a local dev server to spin up ClickHouse + Redpanda + Debezium + Postgres in one command.
Blog is linked in the repo.
Right now the demo solely focuses on the local dev experience, looking for community input regarding best practices for running Debezium in production.
oatsandsugar•3mo ago