Upgrade to Pro — share decks privately, control downloads, hide ads and more …

The first decade of DevOps is over! (TechTalk)

The first decade of DevOps is over! (TechTalk)

Die DevOps-Bewegung wird zehn Jahre alt. Von Beginn an ging es darum, die Kollaboration zwischen Entwicklung (Dev) und Betrieb (Ops) zu stärken. Schon in puncto Tooling ist hier in dieser Zeit viel passiert: Container, Microservices, Public Cloud Services, Serverless u.ä.
Viel tiefgreifender sind aber die kulturellen Veränderungen, die mit neuen Kollaborationsmodellen wie SRE Silos aufbrechen, und neue Methoden wie Continuous Delivery.
Konstantin zeigt, was sich in den letzten zehn Jahren alles im DevOps-Universum getan hat.


Konstantin Diener

June 25, 2020


  1. The first decade of DevOps is over! ein Rück- und

    Ausblick Konstantin Diener | cosee GmbH konstantin.diener@cosee.biz | @onkelkodi
  2. Zoom: https://bit.ly/3hXTEyx

  3. KONSTANTIN DIENER CTO and founder of cosee

  4. Discovery- Phase Backlog Expert-Teams Billing Models Iterative Delivery So ware-

  5. None
  6. 2009

  7. https://de.slideshare.net/jallspaw/10-deploys-per-day-dev-and-ops-cooperation-at-flickr/

  8. https://de.slideshare.net/jallspaw/10-deploys-per-day-dev-and-ops-cooperation-at-flickr/

  9. https://de.slideshare.net/jallspaw/10-deploys-per-day-dev-and-ops-cooperation-at-flickr/ Ops’ job is to enable the business (this is

    dev’s job too)
  10. None
  11. What was 2009 like?

  12. None
  13. Ops: Provisioning of Software, Servers & Networks

  14. On-Premise Hosting

  15. Infrastructure is an INVESTMENT

  16. AWS is 3 years old

  17. Agile Manifesto is 8 years old

  18. Agile Delivery

  19. None
  20. None
  21. None
  22. 2020

  23. Technology

  24. smaller deployment units

  25. 2009 2019

  26. Containers

  27. Public Cloud & Managed Services

  28. Servers, Networks etc. Infrastructure as a Service Managed Service Infrastructure

    as a Service Infrastructure as a Service Managed Service Managed Service Managed Service Managed Service Application
  29. 2009 2019

  30. None
  31. Infrastructure as Code

  32. Cattle vs. Pet

  33. 2009 2020 965-01-node-vs-test irene

  34. Infrastructure as Code is important

  35. But … ☝

  36. • #1 Normalize the technology stack • #2 Standardize and

    reduce variability • #3 Expand DevOps practices • #4 Automate infrastructure delivery • #5 Provide self-service capabilities 2018 State of DevOps Report
  37. But DevOps isn’t just IaC

  38. DevOps is collaboration

  39. Werner Vogels, Amazon “You build it, you run it”

  40. Observability

  41. Obstacles for collaboration

  42. #1 Different Goals

  43. • as many Features as possible • quick response •

    Scrum: Shippable Product • stable application • no Incidents or major bugs • no Downtimes >> many small changes >> rare big changes Development Operations
  44. @techDonaldTrump „I will build a great, great wall between dev

    and ops. I will make ops pay for that wall!“
  45. #2 Handovers

  46. »What? 63 hours, just in queue time?«, Wes say, incredulously.

    »That’s impossible!« Patty says with a smirk: »Oh, of course. Because it’s only 30 seconds of typing, right?« The Phoenix Project
  47. New collaboration models

  48. @noidi „Having a dedicated DevOps person who does all the

    DevOpsing is like having a dedicated collaboration person who does all the collaborating.“
  49. Embedding Inspired by devopstopologies.com

  50. Embedding Ops as a Service Inspired by devopstopologies.com

  51. Embedding Ops as a Service Collaboration Inspired by devopstopologies.com

  52. devopstopologies.com

  53. Site Reliability Engineering

  54. Site Reliability Engineering or Ops done by Software Engineers

  55. Failure & Transparency

  56. Werner Vogels, Amazon “Everything fails all the time”

  57. Chaos Engineering

  58. Blameless Postmortems

  59. Sharing is caring

  60. https://blog.github.com/2018-10-30-oct21-post-incident-analysis/

  61. ChatOps Jason Hand

  62. Outlook

  63. Is operating servers your core business?

  64. Servers, Networks etc. Infrastructure as a Service Managed Service Infrastructure

    as a Service Infrastructure as a Service Managed Service Managed Service Managed Service Managed Service Application
  65. • Keep your proprietary code and infrastructure as minimal as

    possible. • Try to automate everything. • Avoid pet infrastructure.
  66. None
  67. • Learn from failure! • Share your learnings (company +

    customers + community)! • Choose a setup for real collaboration. • Build real cross-functional teams (together with business, support, marketing, …)!
  68. But …

  69. Homework 2030: tackle complexity

  70. Zoom: https://bit.ly/3hXTEyx

  71. Nächster Talk … talks.cosee.biz | blog.cosee.biz | @coseeaner | #coseetechtalks

    Reise zum Mittelpunkt der Corona-Warn-App Tom Brell & Björn Petersen | 30.07.2020