Let’s have some fun while we entertain the idea that maybe - just maybe! - a microservice architecture would be a great fit for our (next) project. After all, we're almost past the microservices hype. It's time to prove that this can work!

Starting from the premise that a microservice architecture is only viable if we focus not only on the speed of change, but also improve the safety of change, we can learn a lot from the early adopters who have already scouted large parts of the unknown territory before us. Tools and platforms are maturing pretty fast too, so it’s safe to assume that we are now (almost) past the peak of impediments. Let’s start enjoying those microservices!

Matthias Noback has been building web application since 2003. He is the author of A Year With Symfony and Principles of Package Design. Currently he is working on his newest book Microservices for everyone. He is a regular speaker at conferences and regularly posts on his blog. While always striving for better programming practices in general, he’s taken a special interest in application architecture, Domain-Driven Design, testing, microservices and application integration patterns.


Comments are closed.

Very interesting. Like I say, perhaps adding an extra slide explaining that you need to understand asyn way of working be able to move to µServices could be interesting. Like you say at the beginning, the naming should be change to something else to help people to focus on the content and not on µServices. (perhaps inter-autonomous-(multi-)services?)

Jos at 14:33 on 3 Jun 2017

As this was the first time this talk was done, it was really good.

It's kind of a talk to motivate people to apply a multi-service architecture if it suits your project.
I really appreciate some pointers as when not to use this technique. Lots of talks try to push a technology/paradigm/architecture and don't mention the projects for which you would kinda, better use another maybe more traditional approach.

Some consequences for the way multi-service projects should be developed were discussed (automating everything: testing, deploying, monitoring) and the argument is made that this will make your development process more mature and faster.
So it's nice to see the balance of investment and return, instead of focusing on just the return if you do it right.

The talk is a bit abstract as there are not many tools discussed that can make it easier to move to, or start a project with multi-service architecture in mind.
There are, however, pointers to bigger, more established ways of making up the elements needed to run a successful multi-service application (like containerization and discovering and defining bounded contexts).

So all in all, a good starter for PHP developers who would like to wrap their head around multi-service applications, but certainly not a talk packed full of tips and tricks with certain tools you can download and work with today.

Multi-service applications still seems like pioneering work in the PHP world.