DevOps: the Data and the Myths

Delivered at Red Hat Summit / DevNation
29 June 2016

Transcript

1. DevOps: the Data and the Myths Michael Stahnke Director of Engineering @stahnma
5. Demographics
8. Employees in high-performing organizations are 2.2 times more likely to recommend their organization as a great place to work.
12. MTTR
13. Change failure rate
14. Isaac Newton’s laws of devops
16. Devops is…
17. Working together
18. Shared Pain
23. Cultural movement
24. “It is impossible to exceed expectations.”
29. Stability vs velocity
34. People, process, tools, measurement
35. People
36. People don’t buy what you do, they buy why you do it. Simon Sinek
 http://bit.ly/sinektedtalk
37. care about the business, why?
38. Diversity
39. Align incentives.
40. Pattern 1: 
 Smooth operations • Devs can teach ops about agile development practices and ops can help devs understand how their apps will run in production. • This pattern works really well for smaller organizations, but it’s extremely important in this structure to have a common toolchain between dev and ops to facilitate better collaboration. Dev Ops
41. Pattern 2: 
 Cross-functional team • Consists of devs, testers, ops, product owner, etc. • Focused on delivering a single application • Self-sufficient • Optimized for throughput
42. Pattern 3:
 DevOps team • Consists ideally of devs with systems experience, or sysadmins with programming experience • Focused on automating pain points • Responsible for building a platform that allows devs to self-service • Provides a toolchain to enable devs to build, test and deploy their systems • Coaches other teams Dev Ops DevOps
44. continuous improvement
45. continuous learning
46. collaboration
48. Process
54. Value stream map
56. Just enough process
57. collaboration to build process
58. Optimized for cycle time - iteration
59. Quality
60. Security
63. ITIL
68. Tools
69. DevOps toolchain 69 Version control Configuration
 management Continuous Integration Deployment
 tools Monitoring and
 others
 … and
 others
 …
70. DevOps toolchain 70 Version control Configuration
 management Continuous Integration Deployment
 tools Monitoring and
 others
 … and
 others
 …
71. Version Control
72. Configuration Management
73. Continuous Integration
74. Deployment
75. Monitoring
76. Measurements
79. Why
85. Metrics gamed
89. Failure: 32K per year - 4.5 m/year
91. Quote 1
92. The shortest path to better software Myth #1: 92 There’s no direct customer/business value for adopting DevOps Practices
93. The shortest path to better software Myth #2: 93 There’s no significant return on investment in applying DevOps principles to legacy applications
94. The shortest path to better software Myth #3: 94 DevOps only works with ‘unicorn’ companies and not traditional enterprise businesses like ours
95. The shortest path to better software Myth #4: 95 Improvement via DevOps principles requires spare time and people that we simply don’t have
96. The shortest path to better software Myth #5: 96 We have regulatory and compliance requirements that preclude the adoption of DevOps principles
97. The shortest path to better software Myth #6: 97 We don’t have any problems that adopting DevOps principles and practices would fix Be “good at getting better” Move faster Eliminate the most frustrating parts of work
98. Learning is not compulsory, but neither is survival. Edward W. Deming
 http://bit.ly/deming14pts
100. • Drive from culture and org • Optimize for feedback, and cycle time • Tools model your process • Measure what matters • Iterate (small batch changes
101. Get your copy of the 2016 State of DevOps Report puppet.com/devops-report-2016