5 Comments

Love this, awesome post my friend Saurabh! I also prefer starting monolith first, then scaling up problematic parts with extracting them to microservices, if it is justified.

Expand full comment
author

Thanks Daniel!

Starting with a monolith is always better because then the microservices adoption is hopefully done for the right reasons.

Expand full comment

Strangler Fig approach using Change Data Capture (CDC) and Kafka that's a great solution, @SAURABHDASHORA

Expand full comment

This is pretty advanced tbh! The closest I’ve come to this was creating an integration test that compared the data read from the monolith against the data read from the microservice. 😄

I like the 3 step roadmap for extracting some code into a microservice, and I used the same approach - without the comparison framework and just relying on integration tests.

Appreciate the shoutout, Saurabh!

Expand full comment

Great article, Saurabh! 🙌 And thank you for mentioning my latest article! 🙏

Expand full comment