Microservices and Biological Systems

I can't find it now but there was a presentation by a Googler on the things they had to do to be reliable at their scale that included a bullet list of ~10 things. I remember going down that list and thinking, "Yeah, that's what life does." for most of them. FWIW. (I'll try to find the list and post a link if so.)

``` You should always start with a monolith. Ensure that it is well tested, well designed and developed with several iterations from the core developers. ```

Upgrading services can be a pain, but starting with a monolith is not a requirement. Architecture can happen before any code is written, and it can illustrate a need for SOA (or not). Difficulty in changing your services as your product matures or as requirements change will be hit at the beginning, but there is not hard stop to that either. Vague statements like the above leave us in the monolith culture indefinitely. It does, however, speak to the need to understand/form patterns for service upgrades.

Don't think of it as your business "failing" because you pursued microservices. You are just participating in an evolutionary process. It's all good.