Topics
1. Five Whys
2. Metrics and Analytics
3. Pick a Boring Tech
4. Devops as a culture
Slide 3
Slide 3 text
Five Whys
Simplest tool for root-cause analysis
Useful for management roles
Useful for postmortem / post-hoc explanations
Pitfall:
1. Sometimes “how” is better
2. It might lead to blaming people
Slide 4
Slide 4 text
Metrics / Data Driven
Principles:
1. Assign owners
2. Create layered metrics, e.g:
● Business metrics
● Development metrics
● Infrastructure metrics
3. Analyze and use for taking decision
Slide 5
Slide 5 text
Metrics / Data Driven(2)
credits: Dan McKinley
Slide 6
Slide 6 text
Metrics / Data Driven(3)
credits: Dan McKinley
Slide 7
Slide 7 text
Pick a boring techs
Innovations tokens concepts
Boring ≠ Bad
Boring means well-understood and lesser
unknown situations / variables
Credit: Dan McKinley
Slide 8
Slide 8 text
Pick a boring techs (2)
Redefine “best tool for the job”
● Our job is “keep our company in business
and make us still eat tomorrow”
● Best tool is “keep problems as minimum as
possible and solve requirement as maximum
as possible”
Credit: Dan McKinley
Slide 9
Slide 9 text
Pick a boring techs (3)
How to pick new techs?
1. Let’s talk first
2. Can we solve with our current techs?
3. Addition or alternatives? Set clear
expectations for migration
Credit: Dan McKinley
Slide 10
Slide 10 text
DevOps as Culture
DevOps is not only tools
Culture movement combined with development
practices to enable rapid development
Slide 11
Slide 11 text
DevOps as Culture(2)
1. Communications
2. Trust
3. Responsibility alignment
4. Respect
5. Send value to customer