With an ever-increasing array of tools and technologies claiming to 'enable DevOps', how do we know which tools to try or to choose? In-house, open source, or commercial? Ruby or shell? Dedicated or plugins? It transpires that highly collaborative practices such as DevOps and Continuous Delivery require new ways of assessing tools and technologies in order to avoid creating new silos. Matthew Skelton shares his recent experience of helping many different organisations to evaluate and select tools to facilitate DevOps; the recommendations may surprise you.

1. How to choose tools for DevOps and Continuous Delivery
Matthew Skelton, Skelton Thatcher Consulting
DevOps Cardiff meetupgroup, 5thNovember 2014, Cardiff, Wales
#devopscardiff

2. Collaboration
Learning
Singleton tools
Conway’s Law

3. Matthew Skelton
•15 years building & operating software systems
•Cybernetics + Neuroscience
•control engineering
•psychology
•‘network’ interactions
@matthewpskelton

4. PIPELINE Conference
London Continuous Delivery meetupgroup (#londoncd)

5. We help you to transform your technology and teams for ‘the cloud’
Interim CTO/Head of X, tech strategy, architecture, workshops, delivery

6. Recent clients
•Legal
•Donations
•Tourism
•Betting & gambling
•Travel booking
•Financial data & reporting
•Healthcare

7. Common themes
•Online since ~1999
•Successful in their sector
•Large, central core database(s)
•Non-aligned goals
•Need to adopt DevOps and Continuous Delivery

8. Common needs
•Technology selection ticklists
•Team interactions
•Collaboration opportunities
•Tools as catalysts

10. Continuous Delivery
•A scientific approach to changing software systems (Dave Farley)
•Regular, rapid, repeatable, reliable changes

11. Continuous Delivery
•Quality
•Safety
•Reliability
•Psychology
•Effectiveness

13. DevOps
“Highly effective, daily collaboration between software developers and IT operations people to produce relevant, working systems” *
*also QA/Testing, IT Service Desk, Programme Management, Commercial, Marketing, etc.

14. Not DevOps
“Automation”
“Build & Release”
“Infrastructure Development”
“System Administration”

16. Collaboration

17. 1. Version Control

24. 2. Deployment Pipeline

25. A large online retailer
•Travel sector
•Since ~1999
•UK market
•‘Non-core’ applications

26. Challenges were:
•Limited Git skills in Service team
•Manual deployments
•‘Snowflake’ servers
•No CI
•Risks: security, automation, etc

27. http://bit.ly/thetrainline-weekly-deploy

28. What we did
•Built a walking skeleton pipeline
•Modelled security roles and stages
•Included manual steps (at first)
•Walked people through steps
•Finally: opened firewall so everyone could see the UI

29. Results
•Security: happy
•Service team: relieved (& happy)
•Developers: won over (& happy)
•Business: surprised (& happy)

30. 3. Log Aggregation

31. LogStash
•ElasticSearch+ LogStash+ Kibana
•ELK
•In Production, Pre-Prod, Test
•On developer machines!!!

32. http://rashidkpc.github.io/Kibana/images/screenshots/searchss.png

34. Collaboration & tool choice
Value collaboration as a key criterion
Orthogonal to main purpose (?)
“How does [the use of] this tool help people to collaborate?”

35. Learning

40. Learning & tool choice
Bring people with you
Appreciate current skills
Prefer achievable gains now
Avoid fear of too-scary tools

41. Singleton tools
(or the ‘Prize Bull’ approach)

43. Singleton tools
•Special database server
•Costly log aggregation
•Costly monitoring
•Server configuration

44. http://www.walpapershddownload.com/highland-cattle-wallpapers/

45. “Better features”?
Optimise globally across the teams that need to collaborate

46. Singleton tool
Breaks feedback (learning) loop from Production
Makes CI/CD more difficult
Underestimates value of collaboration and learning

47. Conway’s Law

48. Mel Conway, 1968
“organizations which design systems ... are constrained to produce designs which are copies of the communication structures of these organizations”
http://www.melconway.com/Home/Conways_Law.html

50. DevOps Topologies

51. DevOps Topologies

52. DevOps Topologies

53. DevOps Topologies

54. Conway & Tool Choice
See the organisation as a system
Separate tools for separate teams
Shared tools for collaborative teams
http://bit.ly/DevOpsTopologies

55. Conway’s Law
•Allan Kelly -@allankellynet
•https://vimeo.com/channels/londoncd

57. How to choose tools for DevOps
Value collaboration aspects
Avoid a learning mountain: evolve tooling
Avoid Production-only tools
Consider Conway’s Law
(this list is incomplete!)

58. Further reading
Build Quality In
buildqualityin.com
70% of royalties donated to Code Club
Discount for @DevOpsCardiffpeople: http://leanpub.com/buildqualityin/ c/DevOpsCardiff2014
Contributors include:
-James Betteley
-Marc Cluet
-Anna Shipman
-John Clapham

59. Thank you
matthewskelton.net / @matthewpskelton
skeltonthatcher.com
HT: @Squire_Matt, @alan_parkinson