Wait, what? Microservices are dead?

If you happened to visit JCON Europe 2023 or follow me on LinkedIn, you may have seen that I presented there using the title “Is it time already? Microservices are dead; long live Microservices!” I know, I know, this kind of title might feel a bit overused, but it expressed what I wanted to convey: Just like Uncle Dave’s “Agile is dead, long live Agile!” we have succumbed to the “tooling approach.” With Agile this resulted in an explosion of Agile methodologies, sometimes even using “project methodology” as a classification, while the whole idea was to put individuals and interactions over processes and tools. However, “doing Agile” does not make you agile!

The same has been happening to microservices, where people are wondering if they should “use microservices” or even “migrate to a microservices architecture.” This is entirely a form of putting the cart before the horse, which can trip you up if you haven’t got the organization to survive this. I’ve already talked about this several times, but wanted to stress this aspect in a different way: If you are properly Agile, and that definitely goes beyond just IT, you will find that striving for small deployable components is a consequence of striving for “Business Agility.” Lean Software Development, DevOps, and the Agile way of working will all help you reduce the time between “Ideation” and “Production.” Somewhere along the line, however, you may find that the size of your application, or its internal structure for that matter, becomes a limitation! Build times, testing, and deployment complexity all battle for attention, and the only way you can successfully conquer them is with smaller component sizes, less coupling between components, and more automation.

I’ll make sure to post the link to the recording of the talk in Cologne when it becomes available!

This entry was posted in Microservices, Software Architecture and tagged , , , . Bookmark the permalink.

Leave a comment