This document discusses the need for DevOps and speed in organizations. It provides five reasons why organizations are looking at DevOps: 1) businesses want faster delivery of products and services, 2) companies aren't lasting as long, 3) increased competition, 4) consumer demand for immediate access, and 5) enabling technologies. The document advocates for a DevOps approach using CALMS principles of culture, automation, lean, measurement, and sharing to build high-trust cultures and continuous delivery pipelines that turn digital assets into business value at greater speed.
3. “The hierarchical structures and organizational
processes we have used for decades to run and
improve our enterprises are no longer up to the
task of winning in this faster-moving world”
John P. Kotter,
Konosuke Matsushita Professor of Leadership
Emeritus at Harvard Business School
https://hbr.org/2012/11/accelerate
4. WHY #1 – Business want it faster!
Design &
Plan
Code Integrate Test Release Deploy Operate
4
Reduce “Cycle Time”
Idea Business value
“Concept” “Cash”
5. Why #2 – Companies aren’t lasting
5
60 years in ~1960
20 years in ~2015
6. Why #3 - “Hyper-Competition”
• Reduced “Barriers to Entry”
• Internet & Cloud “levelling playing field”
• Switch from Capex to Opex
• Cloud makes it easy to start small and grow but still
address a global market
• Reduction in “information asymmetry”
• Customers are better informed (on price, quality via
price engines & online reviews!)
11. “The solution is a second operating system… that uses an
agile, network-like structure and a very different set of
processes… continually assesses the business, the industry,
and the organization, and reacts with greater agility, speed,
and creativity than the existing one”
John P. Kotter,
Konosuke Matsushita Professor of Leadership
Emeritus at Harvard Business School
https://hbr.org/2012/11/accelerate
17. Key ideas
• Software is a Digital Asset
• But Assets can depreciate…
• Assets in “Inventory” cost £££
• Unreleased software costs you money!
• Your Digital Supply Chain gets your Digital
Assets into the hands of your customers (as
fast as possible, safely and securely)
19. What does “Fast” look like?
300 Deployments / Year
50-60 Deployments / Day
10+ Deployments / Day
Every 11.6 seconds
http://www.slideshare.net/grabnerandi/london-web-perfugperformancefocuseddevopsfeb2014
21. DevOps CALMS Model
• Hearts & Minds
• Embracing ChangeCulture
• Automate everything
• Software, Infrastructure, Manual TasksAutomation
• Focus on producing value for the end-user
• Small batch sizesLean
• Measure everything
• Show the improvementMeasurement
• Share everything
• Open Information, CollaborationSharing
23. Build a High-Trust Culture
“In a LOW-TRUST CULTURE, supervision is associated
with words like control, monitor, hover over, and check
up.
In a HIGH-TRUST CULTURE, people supervise
themselves according to the agreement.
The criteria are clear, the consequences are set.
There’s common understanding of what’s expected.
A manager, leader, or parent becomes a source of
help”
– Stephen R. Covey – “First Things First”
24.
25. THE INFLUENCE OF ORGANIZATIONAL STRUCTURE ON SOFTWARE QUALITY
http://research.microsoft.com/pubs/70535/tr-2008-11.pdf
32. Top 5 Predictors of IT Performance
#3 - Proactive monitoring.
“Teams that practice proactive monitoring are able to diagnose and
solve problems faster, and have a high degree of accountability. When
failures are primarily reported by an external source, such as the
network operations center (NOC) — or worse, by customers — IT
performance suffers”
https://puppetlabs.com/sites/default/files/2014-state-of-devops-report.pdf
33.
34.
35. Think Business Value not just monitoring!
Financial
To succeed
financially,
how should
we appear
to our share-
holders?
Objectives Measures Targets Initiatives
Internal Processes
To satisfy
our share-
holders and
customers,
what business
process must
we excel at?
Objectives Measures Targets Initiatives
Customer
To achieve
our vision,
how should
we appear to
our
customers?
Objectives Measures Targets Initiatives
Learning and Growth
To achieve our
vision,
how will we
sustain our
ability to
change and
improve?
Objectives Measures Targets Initiatives
37. Match the Medium to the Message
matching the medium to the message (Lengel & Daft, 1988)
Select the correct channel
depending on the type of
message
38. Many ways of Sharing
1. Daily Stand-ups
2. Group & 1:1 Chat e.g Slack, Google Hangouts or Lync
3. Video Chat e.g. Skype, Google Hangouts or Lync
4. Formal Video Conferencing systems e.g. Cisco, Polycom
5. “Radiator Screens”
6. Email
7. Newsletters
8. “Town Hall Meetings”
9. Internal conferences and Communities of Practice
10.Intranets & Wiki’s e.g. Confluence
11.Document Management Repositories
Informal
Formal
Even the Harvard Business Review says the old ways aren’t good enough…
Accelerate! won the 2012 McKinsey Award, recognizing it as the most significant article published in the magazine that year
John P. Kotter is the Konosuke Matsushita Professor of Leadership, Emeritus at Harvard Business School and the author of 17 books, including Leading Change (reissued with a new preface by Harvard Business Review Press, 2012), The Heart of Change (with Dan S. Cohen; reissued by Harvard Business Review Press, 2012), and (with Lorne A. Whitehead) Buy-In: Saving Your Good Idea from Getting Shot Down(Harvard Business Review Press, 2010). He is a cofounder of Kotter International.
Even the Harvard Business Review says the old ways aren’t good enough…
Accelerate! won the 2012 McKinsey Award, recognizing it as the most significant article published in the magazine that year
John P. Kotter is the Konosuke Matsushita Professor of Leadership, Emeritus at Harvard Business School and the author of 17 books, including Leading Change (reissued with a new preface by Harvard Business Review Press, 2012), The Heart of Change (with Dan S. Cohen; reissued by Harvard Business Review Press, 2012), and (with Lorne A. Whitehead) Buy-In: Saving Your Good Idea from Getting Shot Down(Harvard Business Review Press, 2010). He is a cofounder of Kotter International.
Precision = how accurate we were at predicting something was buggy when it was – when we said it was buggy we were right 86.2% of the time
Recall = proportion of the failure-prone binaries that were correctly identified – we found 84% of the duds
The average precision across all 50 random splits was 0.87 or 87% of the predicted failure-prone binaries that were actually failure-prone. Similarly the average recall value was 0.84, i.e. 84% is the proportion of failure-prone binaries that were correctly identified
http://allankelly.blogspot.com/2013/10/beyond-projects-beyond-noprojects.html
1 million
500,000 £75K risk