Dev, Ops, DevOps, Platform Teams – and everybody has it’s own interpretation. But what is meant by platform engineering, actually? What’s the difference to ops? And isn’t that contradictionary to DevOps?
In this talk, Philipp shows Scout24’s journey from separate dev and ops departments to platform teams and the steps in-between. You’ll see how the organisation changed and why this was necessary. On some real-world examples he explains how the platform teams now uses AWS and Docker to enable the feature teams to become faster and more productive.
9. Conway’s Law
"Organizations which design systems ...
are constrained to produce designs
which are copies of the communication structures
of these organizations."
Melvin E. Conway (April 1968)
11. Why did we change it?
● We wanted to release faster and more often
● Different goals led to discussions
○ Ops: Optimize for stability
○ Dev: Optimize for speed
15. ● Not every Dev became a DevOp
● Ops became not part of DevOps
● VMs were still not in control of teams
● Time to market: More speed by leveraging managed services
Why did we change it?
22. We changed only the 20% rule
● In many teams if felt wrong doing so much infrastructure / shared
services
● Teams wanted to focus on their products
● T-Shape skills often didn't match
● It doesn't scale
Why did we change it?
29. Isn’t that Ops?
NO, because...
● It is a managed product
● Teams can use it or not
● Teams are responsible for their services
30. NO
because...
● It’s not about hosting services from other teams
● It’s about developing products for users
● It’s about products that enable feature teams to work faster
● It’s basically a new mindset
Be careful:
● Do not abstract the cloud
● You’re in competition with the cloud
Is Platform Engineering the new Ops?
31. NO
because...
● DevOps is not a role or team but a culture
● Platform engineering teams should be cross-functional
● Product engineering teams as well
Is Platform Engineering the new DevOps?