DevOps

2 day introduction course

Content

 Day 1
 What is DevOps?
 Dev and Ops
 Which problem does DevOps solve
 Automation

 Day 2
 Performance
 Behavior and Attitude
 Organization
 Governance
 Transformation

2
What is DevOps
Definitions of DevOps
‘DevOps (a portmanteau of "development" and "operations") is a software development
method that stresses communication, collaboration (information sharing and web service
usage), integration, automation and measurement of cooperation between software
developers and other information-technology (IT) professionals. DevOps acknowledges the
interdependence of software development and IT operations. It aims to help an
organization rapidly produce software products and services and to improve operations
performance - quality assurance’ (http://en.wikipedia.org/wiki/DevOps)

Jez Humble
DevOps is “a cross-disciplinary community of practice dedicated to the
study of building, evolving and operating rapidly-changing resilient
systems at scale.”

DevOps is the blending of tasks performed by a company's

application development and systems operations teams.

4
“DevOps is just the Agile principle taken to the full
enterprise”

‘DevOps is just the Agile principle taken to the full enterprise. It provides a more
comprehensive end-to-end perspective that enables enterprise-scale transformation’

‘DevOps is just a natural extension of Agile. It takes the guiding principles and best
practices that Agile brought to developers and applies them on a much broader scale
across the organization.’

* DevOps is Agile for the rest of the company

Mar 4, 2015 Posted by Tony Bradley In Blogs, (http://devops.com/blogs/devops-is-agile-for-the-rest-of-the-company/)

5
What is the current situation? We have built walls –
intentionally or not – that limit performance

Business Development Operations

To help Use
remove DevOps
this wall Here

“DevOps is a response to the growing awareness that there is a disconnect between what is
traditionally considered development activity and what is traditionally considered operations
activity. This disconnect often manifests itself as conflict and inefficiency”
(What is DevOps? Dev2ops.org)

6
The wall of confusion

Development Operations

Extreme focus on change Extreme focus on stability

 More urgent, date/driven projects  Fragile applications failing
 Fragile code going into production  Difficulties identifying root cause
 Releases with ´turbulent installs´  Taking too long to restore service
 Longer release cycles amortize `cost of  Extensive firefighting and unplanned work
deployments`  Unfinished planned project work
 Backlog of infrastructure projects that could fix  Frustrated customers
root cause and reduce costs  Market share decreasing

7
DevOps is ...

DevOps is a way
of building
software while
still getting a
speed to market
without
sacrificing quality
and reliability.

8
What is DevOps

 DevOps is about
identifying bottlenecks
and removing the
waste from the system
to continuously
improve your
organization to deliver
better software.
Bottlenecks can be in
the form of long
approval processes,
manual processes, and
even certain resources.

9
The market sees the solution from various points of
view
What are the points of view of DevOps?

• From functional siloes …

Organization • To multidisciplinary customer-focused teams

• From different KPIs for dev and ops …

Performance • To common goals and measures

• From “different” …
Culture • To “together”

• From complex and slow …

Flow • To end-to-end and simple

• From own tools …

Automation • To integrative tooling

10
DevOps helps to achieve the goals of IT

Quint Definition of DevOps:

DevOps (a portmanteau of development and operations) is an organizational philosophy that
stresses communication, collaboration and integration between information technology (IT)
software developers and IT operations professionals. DevOps uses the interdependence of
software development and IT operations to create higher quality products and services. It aims
to help an organization rapidly deliver the value sought by the customers of IT

DevOps is, in short, “movement” to bring Development and Operations

closer together to achieve four key goals:

• Improve Quality of delivery (First time right)

• Improve Quality of delivery (First time right)
• •Deliver higher
Deliver higherBusiness
Business Value (Customer)
Value (Customer)
• •Increase
Increase Speedofofdelivery
Speed delivery (Performance)
(Performance)
• Improve Efficiency (Reduce Cost)
• Improve Efficiency (Reduce Cost)

11
4.2 New Situation

Making life easier?

Culture Performance

Business DevOps

Process Tooling Organization

4.1
DevOps is the solution to a problem

Automation Performance

Culture

Flow Organisation
DevOps vocabulary

Continuous Delivery is a software strategy that enables organizations to deliver new features to
users as fast and efficiently as possible. The core idea is to create a repeatable, reliable and
incrementally improving process for taking software from concept to customer. The goal is to
enable a constant flow of changes into production via an automated software Continuous
Delivery pipeline.

Continuous integration is the practice, in software engineering, of merging all developer working
copies to a shared mainline several times a day. The main aim of CI is to prevent integration
problems

Test-driven development is a software development process that relies on the repetition of a

very short development cycle: first the developer writes an (initially failing) automated test case
that defines a desired improvement or new function, then produces the minimum amount of
code to pass that test, and finally refactors the new code to acceptable standards.

Test automation is the use of special software (separate from the software being tested) to
control the execution of tests and the comparison of actual outcomes with predicted outcomes

14
DevOps adoption

The adoption of DevOps is being driven by factors such as:

1. Use of agile and other development processes and methodologies
2. Demand for an increased rate of production releases from application and
business stakeholders
3. Wide availability of virtualized and cloud infrastructure from internal and
external providers
4. Increased usage of data center automation and configuration
management tools
5. It has also been suggested that side-effect of the dominant, traditional US
management style (“the Sloan model” vs. “the Toyoda model”)
guarantees the development of silos of automation, thus creating “the
DevOps gap” that then requires DevOps capabilities to address

15
DevOps Goals

 Better align IT responsiveness and capabilities to business

needs
 Produce smaller, more frequent software releases
 Reduce effort with software development, transition and
operation
 Improve time to market
 Improve quality of code
 Improve quality of software deployments
 Reduce cost of product iteration and delays
 Intill a culture of communication and collaboration
 Improve productivity
 Improve visibility into IT requirements and processes

16
Other frameworks
Classic software development

1. Determine requirements

2. Complete the design

3. Do the coding and unit testing

4. Perform functional and

integration tests and fix bugs

5. Perform acceptance and

deploy
Agile Software Development

Most Agile implementations use a routine and formal daily face-to-face communication among
team members. This specifically includes the customer representative and any interested
stakeholders as observers.

Agile development emphasizes working software as the primary measure of progress. This,
combined with the preference for face-to-face communication, produces less written
documentation than other methods.

Agile methods are sometimes characterized as being at the opposite end of the spectrum from
plan-driven or disciplined methods. Agile methods lie on the adaptive side of this continuum.
Adaptive methods focus on adapting quickly to changing realities.

19
Traditional versus Agile

Waterfall Agile
• Start with a complete design • Every Sprint delivers working
• Building is followed by testing the software that can be used in practice
final product • Starts with delivering basic
• Finally, testing in practice functionality to which features are
• No feedback loops added
• Plan-driven • Value-driven
Traditional Approach and No Quality at the Source

21
22
23
Dev and Ops Realities

Dev and Ops are each taking steps to improve their quality and
velocity

 Agile and lean software development practices

 Agile and lean service management practices
 Virtualized and cloud infrastructure from internal and external providers
 Infrastructure as a code
 Data center automation and configuration management tools
 Monitoring and self-healing technologies

Unfortunately, Dev and OPS are not working together on these

initiatives

24
Lean is a paradigm shift

Traditional Lean

Managers have all the answers Manager should ask the right questions,
employees should have the answers as a
team
Managers do the thinking, workers Managers facilitate the workers to add
concentrate on doing value

Activities are done, because they are asked Activities are only done if they add value
to be done for the customer

A certain rate of defects is unavoidable Defects can be eliminated

Knowledge is power Power comes from sharing knowledge and

applying it as a team
 Lean IT, Agile, ITIL and DevOps
• LSD – “Lean Software Development” Application of Lean IT to Software Development – same
principles as Agile.
• “Since one of the Agile principles (and also very Lean) is to make business people and developers
work together daily throughout the project, shouldn’t we call it BusDevOps…or value stream
management
• Although they were independent approaches, they have lots of synergies. A DevOps team could
use:
• Agile approaches to accelerate TTM mainly in Dev side (use SCRUM and XP methodologies),
and involve business (use roles of “product owners” as “business ambassadors”)
• Lean techniques to create and evolve DevOps team both in Dev and Ops sides (Team Vision,
Visual Management, Team Leader Agenda, Retrospective Meeting, VSM, VoC, …)
• ITIL knowledge and best practice to speed up continuous improvement.

How are Lean, Agile, and DevOps related to each other?, Matthias Marschall
27
 Lean, Agile and DevOps – Similarities and
differences

Lean IT Agile/Scrum DevOps

Main Focus Value for Customer Improve TTM Reduce silos inefficiencies

Organization All or a small process Business & Projects Projects & Operations

Visual Mgmt. Core Core Recommended

Project Management
People (B&A) Integral transformation Join Teams (Dev and Ops)
And Devs Teams
Project Management
Processes Value Flow ITSM and PPM processes
(SCRUM, XP, ..)
Tools / CTQ, VSM, Team Vision, User stories, Kanban, Automatic Test. and Deploy.
Artifacts DILO, Kaizen, A3… Product Owner Continues deployment
% Initiatives for improvement % Successful projects % Post-release incidents
% Savings Time & Costs % User stories implemented % Successful changes/releases
Goals / by waste reduction Business satisfaction IT Team performance (E2E)
Indicators Customer and employee Satisf
Quality
28
Lean, Agile and DevOps synergies

• There are synergies in the 3

approaches, although you can follow
your path independently.
• There is no recommended path–
Lean although abroad Lean focus could
IT guide you to cover the most urgent
necessities
• They are no and end by themselves,
but they help you to obtain your goals
and fulfil your requirements (agility,
efficiency, continuous deployment,
continuous delivery, continuous
DevOps Agile / improvement, …)
Scrum • Lean is the most global approach, you
can use this philosophy both for a
particular department or the global
Enterprise

29
DevOps is an evolution

30
DevOps an evolution

31
Evolution DevOps

Design Code Build Integrate Test Release Deploy Operate

DevOps
Continous Deployment
Continous Delivery
Continous Integration
Agile Development

32
Integrate all the useful bits

Tip: View clips at ‘Google Spotify Agile part 1 & 2’

33
Integrate all the useful bits

Tip: View clips at ‘Google Spotify Agile part 1’

34
Integrate all the useful bits (continue…)

Tip: View clips at ‘Google Spotify Agile part 2’

35
Which problem does
DevOps solve
Why did DevOps develop? It is the solution to
one or more problems …

Which issues do we solve?

• Speed of delivery
• Quality of delivery
• Lack of understanding between developers and operations people
• Ability to ensure that customer requests are seamlessly processed
• Bring people who work on the same IT service closer together
• Preventing burnouts

37
38
Importance of Feedback: Three Ways Model

39
The Three Ways

 The First Way – Flow

 Understand and increase the flow of work (left to right)

 The Second Way – Feedback

 Create short feedback loops that enable continuous improvement (right to left)

 The Third Way – Continuous experimentation and learning

 Create a culture that fosters
 Experimentation, taking risks and learning from failure
 Understanding that repetition and practice is the prerequisite to mastery

2 1

40
3
The promise of DevOps
DevOps increases Agility and stability

 Organizations that implement DevOps are more likely to be high performing

 Organisations are more agile
 Code is shipped 30 times faster
 Deployments are completed 8000 times faster
 Services are more reliable
 There are 50% fewer failures
 Service is restored 12 times faster
(source: 2013 State of DevOps Report – Puppet labs and IT revolution press)

42
Same cadence between business and IT

Business processes change over time in a evolutionary way.

So the IT services that enable those processes need to change as well
Who is best

“Amazon has become famous for its high-

velocity deployment of more than 2500
software releases per day across its various
Cloud solution offerings. And there are
similar storied successes at eBay, Etsy,
Facebook, Netflix, Spotify, Twitter.”
http://blog.saugatucktechnology.com/

44
Flow
2 The Flow Solution

The problem is because Dev and Ops use different processes

“The important theme of DevOps is that the

entire development-to-operations lifecycle
must be viewed as one end-to-end process.
Individual methodologies can be followed for
individual segments of that processes (such
as Agile on one end and Visible Ops on the
other), so long as those processes can be
plugged together to form a unified process
(and, in turn, be managed from that unified Key considerations
point-of-view). Much like the question of • Development uses Agile,
measurement and incentives, each
organization will have slightly different Scrum, TDD
requirements for achieving that unified • Operations uses ITIL
process.”
What is DevOps? Dev2ops.org
• Processes are seen to be
different, even incompatible
• Nobody takes an end-to-end
view
3 Key aspects per dimension: Flow

Getting units of work to Flow through the team to deliver value

Flow is about ensuring that units of work move swiftly through the
Flow process
What is it Why is it important
Knowledge of units of The things the team spends improve its performance on each
work time on unit of work
Awareness of How each unit of work needs Developing successful habits
processes to be processed supports team performance

Documentation of How extensively processes are Helps to standardize the way of

processes documented working

Monitoring of progress Understanding the status of Encourages flow efficiency

each unit of work

Customer orientation Each process starts and ends Focuses team on ensuring that
with a customer, and delivers processes deliver value
value
47
DevOps and ITSM

48
DevOps and ITSM (Contd.)

49
Flow for Service Creation and Operation

50
Scrum Approach

51
Scrum Approach (Contd.)

52
Scrum Approach (Contd.)

53
Scrum Approach (Contd.)

54
Scrum Approach (Contd.)

55
Scrum Approach (Contd.)

56
Scrum Approach (Contd.)

57
Scrum Approach (Contd.)

58
 Automation Performance

Continues flow Behavior &

Attitude

Flow Organization

59
 Automation Performance

Continues flow Behavior &

Attitude

Flow Organization

60
Performance
 Automation Performance

Performance Behavior &

Attitude

Flow Organization

The problem is because Dev and Key considerations

Ops measure performance in a • Change vs. Stability
different way • Fast change vs. slow change
“DevOps promises more releases in a • Different KPIs used in development
short timeframe in order to adapt to and operations
problems or changes in the market • Dev and Ops have different goals
faster. […] It also means that the metrics
defined by each team can be
automatically measured and
automatically communicated and shared
with everybody that relies on them to
make decisions.”
Best Practices for Performance-focused
DevOps, ampblog.compuware.com

62
 Automation Performance

Key aspects per dimension Behavior &

Attitude

Flow Organization

Performance is all about understanding how well the team is doing. Delivering value for the
customer; performance must be sustainable.

What is it Why is it important

Key Performance Metrics for understanding Show the team how it is doing.
Indicators performance of the team Identify improvements
Technical Debt Health of the service Directs solutions towards
reduction architectural choices
Level of performance (Inc, Understand performance on Basic hygiene factor for customers.
SR, St. Ch.) small units of work

Level of performance Understanding of performance Steer greatest business value

(creative work) on larger units of work or
specials
Time usage Understanding of how time is Steering the usage of time to
(Earning/Burning) used within the team optimize the time available for
creative work

63
 Automation Performance

Measuring Success Behavior &

Attitude

Flow Organization

Practice that enable organizations to understand and improve flow enable

improved IT performance

 IT performance is measured in terms of throughput and stability

 Throughput is measured by deployment frequency and lead time for
changes
 Stability is measured by mean time to recover and the ability to
preemptively detect and mitigate problems

 Deployment frequency – mean time between deploys

 Cycle time – Time from start of work to ready for delivery
 Change Failure rate
 Mean time to detect incidents (MTTD)
 Mean time to recover (MTTR) – Component
 Mean time to restore (MTRS) - Service

64
Behavior & Attitude
 Automation Performance

Behavior &
Attitude

Culture is all about Flow Organization

Change is good
Communication Tools

Honesty and
Transparency
openness

Rewarding
Collaboration Culture good
behaviours

Grass root
Trust
innovation

Embrace
Accountability
failure
No blame

66
 Automation Performance

Behavior &
Attitude

The culture in Dev and Ops Flow Organization

The problem is a mismatch of cultures Key considerations

between Dev and Ops • Change vs. Stability
“DevOps is simply a culture that brings • Fast change vs. slow change
development and operations teams together • Development people are different
so that through understanding each others’ from Operations people
perspectives and concerns, they can build
and deliver resilient software products that
are production ready, in a timely manner.
DevOps is not NoOps. Nor is it akin to
putting a Dev in Ops clothing. DevOps is
synergistic, rather than cannibalistic.”
Fresh Stats Comparing Traditional IT and
DevOps Oriented Productivity, DevOps.com

67
 Automation Performance

Behavior &
Attitude

Key aspects per dimension Flow Organization

A team of skilled professionals with the right attitude, behavior and motivation to deliver excellent
performance
What is it Why is it important

Team-forming Understanding of where the Helping a team to ‘gel’ to deliver

team is in its evolution high performance

Collaboration An environment in which people This is the basis for continuous

are free to act together improvement and learning

Problem-solving A method and mindset for Ensure progress of and excellence

resolving the issues in the service
Leadership Fit Leadership in relation to the Leader’s style must be compatible
requirements of the team with team needs

Motivation Understanding level of Only motivated people can

motivation of the team produce a sustained excellent
performance

68
 Automation Performance

Behavior &
Attitude

Burn down the old silos… Flow Organization

Burn down the silos and create Stable, co-located teams with a common
integrated cross functional teams goal and enough mandate
 Overcome the challenge of change versus  Product teams and Platform teams
stability  Team KPIs instead of personal KPIs
 Create shared team responsibility for all  Speak the same language
requirements,  Seek first to understand, then to be
 analysis, design, engineering, testing, understood
integration, delivery  Tools (software; Scrum board)
 and deployment  Retrospective and daily stand-up:
 both new development as well as  honest and open feedback
maintenance and support
 Training, knowledge sharing
Results:
 Team building events
 shorter lead time!
 no more walls of confusion
 more trust and fun
 higher quality, lower risk

69
 Automation Performance

… and create transparency Behavior &

Attitude

Flow Organization

Transparency is about getting feedback

 from the customer / business
 from the team
 from the software / system in order to continually improve: inspect and adapt
 Transparency is about visualizing metrics

70
Core of DevOps Culture

71
Cultural Elements of DevOps

72
Changing the Culture

73
Organization
 Automation Performance

Cultural Shift “Connect” your Teams Behavior &

Attitude

Flow Organization

The problem is because Dev and Ops Key considerations

have artificially been organizationally • Traditionally, we have organized skills
kept apart in functional siloes
When we talk about “The Wall” between • Development people sit next to
the Teams (Dev, Test, Ops, Business) – most development people, ops next to ops,
industry professionals we have conversed etc.
with understand and recognize the
concept. So – these walls exist – and we
have to tear them down. The best way to
start is by inviting Testing and Operations
to your daily development Stand-Ups. Let
them listen in what you are currently
working on. Invite them to your sprint
reviews and planning meetings so that they
see what is coming down the pipeline.
Best Practices for Performance-focused DevOps,
ampblog.compuware.com

75
 Automation Performance

Key aspects per dimension Behavior &

Attitude

Flow Organization

Team responsibilities and accountability are, and that it is capable of fulfilling them.

What is it Why is it important

Service Definition Unambiguous definition of the Sense of purpose

service Responsibility and accountability

Customer definition Clear definition of for whom Understanding of value of the

the service is provided service
Completeness of team Skills & knowledge and roles Ensure that the team can take full
to provide the service responsibility for the service

Dependency on other Understanding of shared Optimize freedom to act on behalf

teams resources and dependencies of the customer

Management practices Visual management and other Facilitate team in delivering

governance practices to value
facilitate the team

76
 Automation Performance

Make it easy for customers to do business with you Behavior &

Attitude

Flow Organization
Structure
2 Organizing teams and
the governance

Structure

1
Service
Oriented IT 4
3
(‘DevOps’)
High Performance
Integration High Performing
DevOps Integration High Team model to
Model to guide the Integration Performance guide both
integration of Dev Management Team
and Ops teams and DevOps teams

77
 Define Service-oriented IT based on
Automation Performance

1 Behavior &

customer, service and required skills

Attitude

Flow Organization

The customer IT Service(s)

• Person responsible  All functionality that
for the business supports business
process process
• People using the  Deliver the value
service customers require

DevOps Team
 Responsible for the ‘health’ of a service
 Contains all knowledge and skills required to
support and develop the service
78
Cloud Service

79
Organizational Model: DevOps

Business System Team, manage end user (customer), services and products
(applications)
Platform Team, manage platform product used by Business System Teams

80
Autonomy of Teams

81
 Automation Performance

1 Determining the scope of a DevOps team Behavior &

Attitude

Flow Organization

‘Technology Stack’ of the service

‘Team’ Presentation Layer
Security Layer
SaaS
Desktop Infrastructure ‘Technology stack’
Application Layer determines:
 The area of
Development Platform
responsibility of the
Database PaaS team
Operating System  The required skill set
Server of the team
Storage
Network IaaS

‘Cloud’

A DevOps team is responsible for all facets of:

- The ‘health’ of a service
- The continued match of the service with the business process it supports

82
Focus on Products and Services

83
 Automation Performance

1 Understand demand Behavior &

Attitude

Flow Organization

84
 Continues deployment within the
Automation Performance

1 Behavior &

team
Attitude

Flow Organization

85
 Automation Performance

1
way of working within the teams Behavior &
Attitude

Flow Organization

86
 Create the structure with DevOps
Automation Performance

2 Behavior &

teams and necessary governance

Attitude

Flow Organization

Customers Customers Customers

DevOps Team DevOps Team DevOps Team DevOps Team

Governance

87
 Integrate Dev and Ops, based on a Automation Performance

3 holistic view, starting from the Behavior &

Attitude

current situation Flow Organization

5 DevOps Levels of Integration

Level 1: Low Level 2: Partial Level 3: Initial Level 4: Advanced Level 5: Full
Traditional IT Traditional IT Dev and Ops Dev and Ops DevOps teams
organization organization with closer together together in teams based on a
some DevOps; no based on benefits based on vision integrated vision
clear DevOps of providing value
vision to customers

Automation Flow Organization

Behavior & Attitude Performance

88
 Combine dimensions and integration
Automation Performance

3 Behavior &

levels to achieve DevOps

Attitude

Flow Organization

DevOps is characterized by integration rather than maturity

Level 1 Level 2 Level 3 Level 4 Level 5

Individuals, lack of High-performing integrated
Organization management and team with agreed practices
technical practices

Performance No measures, no Measures integrated into

steering way of working

Behavior & Basic leadership, Facilitating leadership, high

Attitude low motivation motivation and continuous
improvement
No customer Processes are simple and
Flow orientation, poor short, few descriptions
processes
No or little tooling Fully integrated suite of
Automation for testing and tooling for development,
deployment testing, deployment and
service management

89
 Focus on turning the DevOps team Automation Performance

4 into a High Performance Team that Behavior &

Attitude

delivers Customer Value Flow Organization

Team Development
Leadership

High
Vision &
Measurement Flow Performing
Goals
Team

Collaboration

90
Governance
structures
Governance within Teams

92
BI Team use agile techniques to manage
workload

Incidents
Team
Product Backlog
• Intakes
• Projects
BI Governance Team

• Functional
changes
Customer

• Source changes
e

• Problems
• Advice
• Technical debt
• …

Operational activities,
service requests

93
BIGT directly linked to teams

BI Governance Team

Team Team Team Team Team Team

Customer and Asset Execution Optimization Service & Staff Network
Market Management

Architecture & Quality

IT BI Delivery Group

BIGT: Benefits
• All changes according to architecture • Direct steering of delivery
• Checks execution with availability of reduces conflicting changes
resources • Central quality check to avoid
• Changes allocated to most suitable team new technical debt

94
BI Governance Team: Way of Working and
Timing

Week k Week l Week m Week n Week o

Meeting 1 Meeting 2 Meeting 3

BIGT BIGT BIGT

backlog planning execution
IMs & Architects Scrum masters Scrum masters
& Architects

Sprint Sprint Sprint Sprint Sprint

Week 3 Week 1 Week 2 Week 3 Week 1

95
Key part of improving the service: Knowledge
development

Team Team Team Team Team Service Team

Customer & Asset Mgmt Execution Optimization & Staff Network
Market

Knowledge meetings
(min. 1 x / sprint; 1 rep per Team mandatory)

Benefits
• Up-to-date-knowledge
Solution architect Experts discuss • Common technical way
determines • Knowledge & learning of working
• Knowledge and goals • Prevents solutions that
learning goals • Technical do not fit in
• Technical Guidelines coordination architecture

96
Transformation
 Approach creates predictability and stability in
a changing organization and space for
realizing ambitions
Analysis and
Structure Integrate Perform
definition

“Current state?” “ Organizing teams and “Working in DevOps “High performance on

“Future state?” governance” teams” collaboration and technology
“Commitment?” ”
▪ Service Oriented IT analysis ▪ DevOps training ▪ Form DevOps teams ▪ Intensify collaboration
▪ Agreeing design principles for ▪ Detailed team and ▪ Awareness and training ▪ Improve flow
DevOps teams and governance designing ▪ Prepare Team for integration ▪ Extensive automation
governance ▪ Defining technology stacks ▪ Analysis of improvement ▪ Multi-skilled team members
▪ Leadership workshop(s) on ▪ Transformation roadmap potential on DevOps goals ▪ Focus on customer-oriented
DevOps principles, goals and workshops performance
behavior

Results Results Results Results

in in in in

▪ Current state of Service ▪ Detailed definition of ▪ Detailed Integration plan per ▪ Improved Quality of delivery
Oriented IT customers, services, teams team ▪ Higher Business Value
▪ Design principles and technology stacks ▪ Trained employees ▪ Increased Speed of delivery
▪ Commitment leadership to ▪ Roadmap and ▪ Continuous service ▪ Improved Efficiency
DevOps principles and goals communication plan improvement roadmap

▪ Management ▪ Management, Architects, ▪ Management, Teams in ▪ Management,

Lead technologists, Team scope DevOps teams
leads, HR 98
Getting DevOps to work

DevOps is an individual path per team

• Choose a common reference Good practise implementation approach Agile/Scrum maturity level 1

framework (DevOps Integration

Prepare / kick off Sprint 1 (two weeks) Sprint 2 (two weeks) Sprint3 Sprint 4 Sprint 5 Determine Next steps after implementation

Product back log Focus on Done

Plan meetings
Process -planning

Model) to facilitate learning

-daily scrum
-review
-demo
Focus on Ready

Maturity scan Bord 1.0 Bord 2.0 Bord 3.0

Focus on: to do - Focus on: Focus on:
doing - done burn down - velocity
integrating Maturrity scan and
Performance determine next
including Kanban DEV en OPS
steps

•
board

Assess the current situation per Identify and train:

Product owner
Scrum master
Setup Team
Skills matrix clear Involve
users in the
Involve
chain

team/department Organization
The Why of the
processes and use
of agile and devops

Understand Create Team Workshop

5 cornerstomes Charter Determine Team
Behavior & dynamics

•
Attitude

Create individual plan per team to Identify Application

achieve the selected level of

Landscape & Value Vision & Plan for
Vision & Plan for implement
Chains Automation
Config and automation
Tooling
Identify used standards

integration Town Hall

Agile Awareness
in training
determine
first DoD and Deep dive on Deep dive on Code Deepdive on Scaled
The Skills DoR def Story mapping Quality Agile Framework
(coaching) first sprint
Scrum Foundation Other necessary
backlog
deepdives
first spint poker Deep dive on kpi's and Other necessary
Train SM & PO etc. metrics deepdives

• Pay attention to the governance of

Change Story

Other necessary Coach the coache

Lean & Agile Other necessary approach
The Leader Visit other team & deepdives deepdives
Champion Leadership
(coaching)

the DevOps teams

99
Start with assessing the current situation of
the team, using the Quint DevOps Integration
Model
The Integration model from Dev and Ops to a DevOps team, and on to a high-performing
DevOps team

DevOps Integration
Low Partial Initial Advanced Full
Level 0 Level A Level B Level C Level D Level E
Organization Service Definition 1 2 3 4 5
Customer definition 1 2 3 4 5
Completeness of team 1 2 3 4 5
Dependency on other teams 1 2 3 4 5
Management practices 1 2 3 4 5

Performance Key Performance Indicators 1 2 3 4 5

Technical Debt reduction 1 2 3 4 5
Level of performance (Incidents, SRs, St Ch.) 1 2 3 4 5
Level of performance (Creative work) 1 2 3 4 5
Time usage (Earning/Burning) 1 2 3 4 5

Behavior & Attitude Team-forming 1 2 3 4 5

Collaboration 1 2 3 4 5
Problem-solving 1 2 3 4 5
Leadership fit 1 2 3 4 5
Motivation 1 2 3 4 5

Flow Knowledge of units of work 1 2 3 4 5

Awareness of processes 1 2 3 4 5
Documentation of processes 1 2 3 4 5
Monitoring of progress 1 2 3 4 5
Customer orientation 1 2 3 4 5

Automation Automated Test 1 2 3 4 5

Development Tooling 1 2 3 4 5
Automated Deployment to Production 1 2 3 4 5
Service Management Tooling 1 2 3 4 5
Use of other tools 1 2 3 4 5

10
Integrate Technical and Management Methods

Ensuring the right habits makes the DevOps team more effective

• Standardize methods to encourage learning

• Define the interface with the customer,

together with the customer
Agile & Continuous Delivery • Define the flow of units of work through
the team, e.g. Waterfall, DSDM, Agile,

• Define the management principles used by

the team, e.g. Lean, TOC,

Lean & Visual Management

101
Success factors

 Start small with quick wins

 Attention to culture and tools
 Transparency and adoption
 Ensure the process is end to end
 The attitude of ‘doing it right now’
 Automation of testing and deployment
 Track and analyze impact of changes
 Constant cooperation within the teams and over the teams
 Get buy in from senior management

102
Big Bang VS Snowball

Big Bang Snowball

Disadvantages Disadvantages
 Critical processes can’t be tested first  Number of people are first and some are last,
 Changes apply to all, could be difficult to master jealousy my occur.
and the impact could be huge  Requires planning.
 No flexibility in scheduling
 Impact of changes are difficult to measure, it all Advantages
comes at once  Changes only apply to small groups, impact is
small
Advantages  Incidents can be sorted out and conquered first, to
 It applies to everyone, no jealousy stabilize before taking the next step
 Requires little or no planning.  The process are tested frequently and become
more and more reliable.
 The process is reviewed over and over,
inefficiencies could be optimized every time.
 Small changes early results

103
Automation
“We find ourselves now with two camps in the DevOps
community: technical and non-technical. Either side shifts the
conversation to where they feel most comfortable: the domain
they work in and understand. Technical folks talk about tooling.
Non-technical folks talk about culture.”
http://dustinrcollins.com/devops-a-house-divided

10
Continuous delivery covers the full
deployment pipeline

Design Code Build Integrate Test Release Deploy Operate

 Treating infrastructure as code by programmatically provisioning and

managing infrastructure resources
 Repeatable and reliable deployment processes
 Development and testing (preferably automated testing) performed against
production-like systems
 On-demand creation of development, test, staging and production
environments
 Proactive monitoring of infrastructure components, environments, systems
and services

10
ENGINEERING FLOW OF CONTINUOUS DELIVERY

Design Code Build Integrate Test Release Deploy Operate

Check in Trigger

Feedback

Check in Trigger

Feedback

Check in Trigger
Trigger Trigger
Trigger

Feedback

Check in
Trigger
Trigger Trigger
Trigger
Trigger

Feedback

Check in Trigger
Trigger Trigger
Trigger
Trigger Trigger
Feedback

10
Continuous Integration, delivery &
Automation Performance

Behavior &

deploy
Attitude

Flow Organization

Continuous Integration: Integrate early and often to minimize the gap between development
and operations to increase:
 quality assurance
 test driven development
 configuration management
Continuous Delivery: extension of continues integration and ensures that every change is
releasable:
 Automation to improve processes software delivery including integration and acceptance
testing.
 Release in one push on the button
Continuous Deployment: end to end software delivery which includes continues integration and
delivery:
 Automate the release of the delivered software
 Automated infrastructure, Release Management
 reduced lead time, earlier feedback, reduced risk

Design Code Build Integrate Deploy Test Release Operate

Continuous Integration Continuous Delivery Continuous Deployment

108
 Automation is vital for ensuring speed,
Automation Performance

Behavior &

consistency and quality of work

Attitude

Flow Organization

What is it Why is it important

Service Management Tool in which all units of work Monitoring progress and data for
Tooling are registered the overall KPIs
Automated Test Tooling to ensure the quality, A tool can generally test more
completeness and speed of quickly and accurately
the test process.
Automated Deployment to Automated flow from dev Reduces the chance of mistakes;
Production through test and acceptance ensures the consistency of
to production environments
Development Tooling Tooling supporting the Reduces the time required for
development process developing working software

Use of other tools Non-technical tools e.g. Not everything is integrated

checklists, guidelines and into tooling, but does need to
architecture drawings. be available for the team

109
Continues Delivery Implies: Software has to Flow

110
Continues Delivery Implies: Software has to Flow
(Contd.)

111
Type of Feedback

112
TYPICALTOOLS THAT SUPPORT THIS PRINCIPLE OF
CONTINUOUS DELIVERY

Contin.
ALM SCM Build Analyze Reposit. Test Deploy Monitor
Integr.

Nexus

Endevour Endevour
ITCAM

TADDM IWD TBSM

Endevour

113
 Automation Performance

Tools for automation Behavior &

Attitude

Flow Organization

114
 Automation Performance

Behavior &
Attitude

Benefits of automation Flow Organization

Automation supports Automation enables

 Faster lead times computers to do rote tasks
 More frequent releases and allows people to
 Less turbulent releases  Solve problems
 Fewer errors  Make decisions based on feedback
 Higher quality  Use their skills, experience and
 Faster recovery Judgment
 Business and customer satisfaction

115
Considerations
Considerations with DevOps

 Legacy systems: first do DevOps then let the team clean up the legacy
 Team forming
 Short cycles of feedback from customer to DevOps team
 Short cycles of feedback from P to DTA
 Real service definition
 Encourage the self-cleansing capability of the team (“Eat your own dog food”)
 Sizing for DevOps – how big must my organization be?
 Shared resources or underutilized resources
 What happens to all the process management roles?

117
 Other considerations … (1)
4.2

There are other details to be ironed out

How to deal with the old-school segregation of duties?
• The release manager role ensures that there is no unauthorized access to the
production environment.
• As a result of the increase in deployments, the need for quick, unauthorized
access is reduced as the process fulfills the needs of the team.
• Tooling can also ensure that SoD is facilitated

How to deal with the ‘shared resources’ phenomenon?

• “I have one DB expert who reviews all major queries on our database system.
But now I have 4 DevOps teams? In which one should he be?”
• The key is to understand how much time the expert needs to spend per team.
• If, for example, there is 1 day of work per team, then the shared resource
should focus on sharing knowledge and skills with people in the team. The
expert can then be used for the specialist tasks and project support.
• This expert should be evaluated on flow efficiency rather than resource
efficiency.
Other considerations … (2)
There are other details to be ironed out
How to deal with the promise of increased efficiency?
• Since I get faster, automated and better at things…, I guess this whole
DevOps thing is a great way as a drastic cost saver, no?
• There is a basic minimum size to a DevOps team; below this level, the service
can no longer be guaranteed. The aim is to continually increase the amount of
value added by the team. The IT service can always be further automated and
improved.
• If there is truly too little work, expand the scope of the team.

How to deal with customers?

• “Since one of the Agile principles (and also very Lean) is to make business
people and developers work together daily throughout the project, shouldn’t
we call it BusDevOps…or value stream management?
• The intention of DevOps is to improve the flow efficiency of the units of work
that enter the team. Involving the customer gives a greater certainty that the
team is also dealing with the right things
• A DevOps team must have its processes in place so that the customer is
enticed to join in
Other considerations … (3)
There are other details to be ironed out
How to deal with visualization?
• My Agile board starts with epics, to features, to sprint items, to stuff in integration
and then ‘done’…; so what columns do I add once Ops is in my team? Are we still
working on the same units of work?
• The units of work are identical only we use generic common names for them
(incident, change, project, service request, problem). These are used on the team’s
VM boards. The details per unit of work will differ

How to deal with governance?

• “I’m currently a project manager. I already have two scrum masters on board of my
development team plus a product owner. And now there’s this team leader of the
Ops team. Who is going to be the appointed manager in this circus and what is its
main responsibility?
• It works the other way round. First, there is a service. Second, there is a team that
supports it. Within the team, all units of work are processed. The ‘project manager’
may have the role of team leader, or someone else may have the role. In this case,
the project manager reports to the team leader.
Other considerations … (4)

There are other details to be ironed out

How to deal with ‘interest’ or lack thereof?
• One of the traditional arguments against DevOps is that the people involved are not
interested in each other’s work: a developer does not understand and is not
interested in operations and vice versa
• This is the cultural change that needs to take place in the team: there is a common
goal – the health of the IT service and everyone plays an equally important part. It is
vital for the team spirit that each team member is aware of the contribution that
each of the team members makes.
• The lack of interest is something that we have created as a result of out way of
steering and structuring the organization and individuals
You’ve been introduced to DevOps

Now what?

122