niggl/cnsmunich25
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
cnsmunich25/content/day2/07_devex.md
Nicolai Ort b4b5c11f12
All checks were successful
Build latest image / build-container (push) Successful in 45s
Details
fix: ":" in title
2025-07-22 15:55:01 +02:00

2.5 KiB
Raw Blame History

title, weight, tags
title weight tags
Creating a smooth Developer Experience: from complexity to simplicity 7
dx
devex

Why?

  • Complexity in software fev is increasing
  • Kubernetes itself is simple, but the landscape is gigantic
  • Automation reduces complexity, but creating automation adds complexity

Cognitive Load

  • Instrinsic Load: Natural complexity of a task
  • Germane: Effort needed to build expertise and optimise workflows
  • Extraneous: Unnecessary load caused by poor tools or processes combined with distractions

Example for the rest of the talk

  1. Dev: Needs Environment
  2. Ops: Sets up the env in multiple steps
  3. Dev: Receivs new env

Problems

  • Maybe the ops person has no time
  • Doing things via tickets takes a shit ton of time

Our Lord and savior: Platform Engineering

  1. Centralize
  2. Automate: Automate the devops cycle
  3. Simplify: Solve complexities associated with the cloud
  4. Efficency: Avoid bottlenecks
  5. Quality and Reliability
  6. Self-Service

Circular Economy

  1. When a team creates a service, it can it to a marketplace
  2. Another team cann use it, kustomizue it and publish the customized version to the marketplace

Metrics

  • The invisible engine behind every innovation -> No improvement with out proving it
  • Who wants to know what?
    • User: Wants to measure the platforms impact on dx and productivity (Feedback, Provisioning time, ...)
    • Org: How does the platform support org efficiency and delivery speed (Time to market, cost savings, ...)
    • Platform: Stability and team effectiveness (Mean Time to Resolution, uptime, ...)
  • Metric-Frameworks: DORA (DevOps Research and Assesment), SPACE (Satisfaction Performance, Activity, Communication and Efficiency)

And how do we drag AI into this?

  • Frameworks kann make dx measureable
  • Challenge: NEw tools and processes can create friction
  • AI Can help with: Assistance, Code generation, Debugging, Conversational explaination instead of reading docs yourself

Tips

  • Adopt the right mindset (Why a platform, what do i need and how would it impact exsisting metrics)
  • Invest in a effortles selfservice experience
  • Gradualy move habits towards the platform - don't force everything at once
  • Create solutions that people enjoy using