Continuous Delivery • A GitHub Story

Transcript

1. How I learned to stop worrying and love the deploy CONTINUOUS D E L I V E R Y
2. matthew@github.com github.com/training @matthewmccull matthewmccullough
5. Why? The current state of affairs
10. we have a problem
12. "This is so important, we can't entrust it to automation. We need our top developer to supervise it."
13. The Loose Parts Building blocks of better results
14. automation everywhere
15. continuous integration
16. agile practices
17. traceability
18. rollback
19. toggles
20. humane interface
21. Divide and Conquer Small actions are generally best
25. Automate Humans are creative. Robots are consistent.
28. continuous integration servers
29. continuous integration servers jenkins travis ci circleci buildhive
35. Embrace Transient-ness Hardware on demand provides flexibility
36. Azure
38. AppHarbor
40. Cloudbees
42. Amazon Web Services
44. Heroku
46. Hubot
48. hubot: <dosomething>
49. Measure Identify mistakes at home and in the wild
58. Feature Flags Dip toes, not heads, into unknown waters
59. http://www.flickr.com/photos/mwichary/3209181446/
62. Fail Forward Fix the root of problems
63. git revert
64. small changes
65. try not to roll back
66. fail forward
67. Listen to Leaders Open sourcing all of the ideas
71. Risks What's the worst that could happen?
72. kill some features
73. discover we're wrong
74. cause some regressions
75. Q&A
76. matthew@github.com github.com/training @matthewmccull matthewmccullough