The road to Continuous Deployment

Deploying code changes continuously to your production environment has many benefits. New features get to your users much faster which enables very short and quick iterations on them. And due to the small change sets you tend to encounter less bugs and regressions. In this talk I’ll present what Continuous Deployment means, and what you have to do to make it work successfully. We will cover topics like trunk based development, feature flags, partial roll outs, continuous integration, deployment automation and application monitoring.

Transcript

1. The road to Continuous Deployment @BastianHofmann
2. What?
3. Why?
4. How?
5. A few words about me
13. Questions? Ask
14. http://speakerdeck.com/u/bastianhofmann
15. ?
16. Company?
17. Freelancing?
18. Single product?
19. Agency?
20. Do you know … / Do you do …
21. Continuous Integration?
22. Continuous Deployment?
23. Feature flags?
24. Central log management?
25. KPIs?
26. To the glass monitoring?
27. If not, you will after the talk.
28. So
29. Continuous Deployment
30. Deploying code changes to production continuously
31. Classic approach
32. Waterfall
33. http://en.wikipedia.org/wiki/File:Waterfall_model.svg
34. Web! Hello?
35. Agile
36. http://en.wikipedia.org/wiki/File:Scrum_process.svg
37. Going further
38. Continuous Deployment
39. Benefits
40. Smaller change sets
41. Less Bugs
42. Finding the bug
43. Shortest „time to bugfix“ possible
44. Better performance
45. Better conversion
46. Faster iterations
47. Awesome, I’m sold
48. How can I introduce it?
49. The whole team has to live it
50. Not only doing more releases!
51. Step by step
52. Version control
53. is very important
54. Manual QA
55. does not scale
56. Automated Tests
57. Unit Tests
58. Integration Tests
59. Acceptance Tests
60. Smoke Tests
61. JS Tests
62. Regression Tests
63. Continuous Integration
64. Jenkins http://jenkins-ci.org/
66. Development process
67. Feature branches are bad
70. Merge Hell
71. -> Big changes
72. Trunk/master based development
73. Every commit gets integrated
74. Every commit gets deployed
75. No commits that are not stable
76. Frequent release branches with cherry pics
77. http://paulhammant.com/2013/03/04/facebook-tbd/
78. http://paulhammant.com/2013/05/06/googles-scaled-trunk-based-development/
79. http://paulhammant.com/2013/03/04/facebook-tbd/
80. But there is stuff that takes longer then an hour to develop
81. Feature Flags/ Toggles
82. Release !== Deployment
83. public function hasAccess(! $accountId, array $roleNames! ) {! return featureFlag()->isActive(! FeatureFlag::TEST_ONE! );! }
86. Partial rollout
87. Experiments (A/B Testing)
88. other also possible: goal early integration
89. Finally: Getting your code live
90. Automatic builds
91. Composer
92. Bower
93. NPM
94. Browserify
95. JS
96. CSS
97. HTML
98. Static caches
99. Ant
100. Phing
101. Grunt
103. Custom scripts
104. Fast builds
105. Hook it into Jenkins
106. Automatic deployment
107. One-click- deployment
108. Fast deployment
109. Fast rollbacks
110. http://www.ansible.com/
111. http://capistranorb.com/
112. Etsy Deploynator http://codeascraft.com/2010/05/20/quantum-of-deployment/
113. Custom scripts
114. Cache warmup
115. Asset invalidation
116. https://www.npmjs.org/package/grunt-cache-invalidate https://www.npmjs.org/package/gulp-rev https://www.npmjs.org/package/grunt-static-ctrl
117. Staging System(s)
118. Central DEV Production Near Production Integration
119. Build/Test/Release pipeline
121. Schema migrations
123. Non BC breaking
124. Only safe operations
125. Migration in steps
126. No software is bug free
127. Finding bugs as early as possible
128. Detecting errors
130. Central log management
131. More this afternoon
132. Detecting regressions
133. Bugs
134. Performance
135. KPIs
136. Plotting everything
137. Graphite http://graphite.wikidot.com/
139. More this afternoon
140. Finding the cause
141. Anomaly detection
142. http://codeascraft.com/2013/06/11/introducing-kale/
143. Profiling
144. XHProf https://github.com/facebook/xhprof
145. Use it in production for a subset of requests
146. XHGUI https://github.com/perftools/xhgui
149. Qafoo Profiler https://profiler.qafoolabs.com/
151. New Relic http://newrelic.com/
153. Symfony Debug Toolbar
156. Extend it http://symfony.com/doc/current/cookbook/profiler/ data_collector.html
159. Aggregate it
161. Synopsis
162. Invest in your processes
163. Invest in quality
164. Invest in performance
165. Invest in agility
166. Continuous Deployment
167. http://twitter.com/BastianHofmann http://lanyrd.com/people/BastianHofmann http://speakerdeck.com/u/bastianhofmann ! mail@bastianhofmann.de