Even as a proponent of what I perceive to be the core concepts of microservices, I think building hundreds of very small ones is a bad idea

Dec 3, 2016 · 12:11 PM UTC

8
31
1
44
Replying to @stilkov
I'm seeing the move to Lambda/FaaS break down services to collections of much smaller functions. Seems to be productive so far.
2
1
3
Definitely the future model in terms of infrastructure support, I won’t disagree. I still think a larger “unit” is needed, too
4
more replies
Replying to @pavlobaron @adrianco
Despite being a cool platform, Lambda is just one potential choice. How will I move to the next thing in five years?
Replying to @pavlobaron @adrianco
The point, to me, is to allow for independent evolution of parts
Replying to @stilkov
What about in the context of a fabric (like #ServiceFabric or Akka)?
Replying to @pavlobaron @stilkov
the microservices model supports independent evolution of all the components. Lowers adoption and migration costs.
1
Replying to @stilkov
@raymcdermott we had a tool long before discussing microservices: DDD and Bounded Context. Your answer to size, not only technical.
1
Replying to @pavlobaron @adrianco @stilkov
Size is about the expressiveness of the PL (think APL) == unimportant. MS "size" is about responsibility IMO.
1
Replying to @stilkov
how small is "very small" from your perspective?
Replying to @stilkov
So microservices can scale but the idea doesn't?
Load more