This document discusses several topics related to DevOps. It begins by noting that DevOps is becoming a more mainstream part of IT strategies, though more "de-training" of CIOs is still needed. It also discusses the debates around DevOps certification and incentives. Two experts provide opinions on microservices and data consistency. The document advocates for training programs to develop DevOps skills at multiple levels and notes that not all DevOps roles require "rock stars" but rather different calibers of individuals. Overall it addresses challenges in DevOps recruitment, skills gaps, and changing perceptions and job titles in the industry.
This document discusses various perspectives on DevOps. It begins by noting that DevOps adoption is increasing in the enterprise and that key requirements include proper incentivization of employees. It also discusses debates around DevOps certification and the increasing focus on microservices. Several articles then provide opinions on whether DevOps is too elitist, the challenges in DevOps recruitment, and the need for DevOps training programs to address skills shortages. The document advocates treating DevOps practitioners with respect and focusing on hiring solid developers rather than "unicorns" or "rock stars".
This document discusses different perspectives on whether DevOps is too elitist. Some key points made include:
- DevOps demands close collaboration between development and operations teams, challenging traditional roles, which some see as elitist.
- High salaries and demand for DevOps skills could breed a sense of superiority, but this is normal for cutting-edge fields.
- There is a need for more training to develop DevOps skills across different experience levels to address the skills shortage.
- Focusing too much on "rock star" candidates or buzzword skills could cause some organizations to miss out on highly experienced people.
This document discusses DevOps and the use of metrics and data in DevOps. It provides perspectives from DevOps leaders on some of the challenges around collecting and using metrics and data in a DevOps environment. Some of the key points made include:
- Collecting too many metrics can create "noise" that obscures useful metrics. Metrics should either inform teams about issues or validate hypotheses.
- Useful metrics indicate anomalies or trends that need investigation and resolution, or provide proof against a hypothesis.
- When determining metrics, understand what information is needed, who needs it, and why, to ensure the metrics will be actionable.
- The volume, variety and velocity of big data can be achieved in Dev
Keeping your career secure presentation august 2013Fernando Herrera
As you get older, you get slower, uglier and more expensive. As your work rate goes down and your cost goes up, you must compensate by expanding your skills and adding more value to the organization. Whether you are starting out on a career in IT or on the home stretch to exiting one, this paper looks at the threats to your career security and the strategies for ensuring you stay in one.
The document discusses several topics:
1) It asks whether employers still expect to see CVs in addition to online application forms, and finds that many employers, especially smaller companies and those in certain sectors, still prefer and request CVs.
2) It provides feedback from a workshop on preparing PGCE applications, finding that most students found the content and exercises useful.
3) It announces an upcoming talk about the HMRC graduate recruitment scheme.
Trilogy University (TU) is Trilogy's corporate training program for new hires modeled after the US Marine Corps boot camp. It pushes recruits to their limits through intense challenges and teamwork to solve real business problems under tight deadlines. Graduates emerge highly confident and bonded to each other and the company.
TU achieves more than typical corporate boot camps by serving as Trilogy's primary research and development engine and incubator for strategic thinking and next generation leadership. This is possible because Trilogy's top leadership is deeply engaged in TU, participating directly and using it to develop the company's future leaders. The program has become a virtuous cycle that continuously improves due to this leadership involvement and commitment to innovation.
The Extreme BlueTM internship program at IBM places top students on project teams to develop new products and services. Interns submit many patent disclosures and open source contributions. They help create solutions for clients and bring new IBM products to market. Unlike other internships, interns work on leading technologies like cloud computing and petabyte-scale data analysis.
The AT&T MATREX IT Internship offers interns experience in the inner workings of a technology company through challenging assignments, exposure to technology teams, and hands-on experience.
Cisco offers flexible internships globally in various business departments, from 3-18 months. Internships provide opportunities in fields like sales, engineering, finance, and marketing.
Sabine Douglas is a software and mobile development consultant who has extensive experience recruiting for these fields. She focuses on recruiting candidates with specific technical skills rather than targeting particular industries. Sabine works with organizations of all sizes, from start-ups to large companies. She has a strong presence in the North West of England but finds mobile development candidates across the UK. Recommendations praise Sabine for her specialized expertise, high-quality placements, and personal approach to understanding clients' and candidates' needs.
This document discusses various perspectives on DevOps. It begins by noting that DevOps adoption is increasing in the enterprise and that key requirements include proper incentivization of employees. It also discusses debates around DevOps certification and the increasing focus on microservices. Several articles then provide opinions on whether DevOps is too elitist, the challenges in DevOps recruitment, and the need for DevOps training programs to address skills shortages. The document advocates treating DevOps practitioners with respect and focusing on hiring solid developers rather than "unicorns" or "rock stars".
This document discusses different perspectives on whether DevOps is too elitist. Some key points made include:
- DevOps demands close collaboration between development and operations teams, challenging traditional roles, which some see as elitist.
- High salaries and demand for DevOps skills could breed a sense of superiority, but this is normal for cutting-edge fields.
- There is a need for more training to develop DevOps skills across different experience levels to address the skills shortage.
- Focusing too much on "rock star" candidates or buzzword skills could cause some organizations to miss out on highly experienced people.
This document discusses DevOps and the use of metrics and data in DevOps. It provides perspectives from DevOps leaders on some of the challenges around collecting and using metrics and data in a DevOps environment. Some of the key points made include:
- Collecting too many metrics can create "noise" that obscures useful metrics. Metrics should either inform teams about issues or validate hypotheses.
- Useful metrics indicate anomalies or trends that need investigation and resolution, or provide proof against a hypothesis.
- When determining metrics, understand what information is needed, who needs it, and why, to ensure the metrics will be actionable.
- The volume, variety and velocity of big data can be achieved in Dev
Keeping your career secure presentation august 2013Fernando Herrera
As you get older, you get slower, uglier and more expensive. As your work rate goes down and your cost goes up, you must compensate by expanding your skills and adding more value to the organization. Whether you are starting out on a career in IT or on the home stretch to exiting one, this paper looks at the threats to your career security and the strategies for ensuring you stay in one.
The document discusses several topics:
1) It asks whether employers still expect to see CVs in addition to online application forms, and finds that many employers, especially smaller companies and those in certain sectors, still prefer and request CVs.
2) It provides feedback from a workshop on preparing PGCE applications, finding that most students found the content and exercises useful.
3) It announces an upcoming talk about the HMRC graduate recruitment scheme.
Trilogy University (TU) is Trilogy's corporate training program for new hires modeled after the US Marine Corps boot camp. It pushes recruits to their limits through intense challenges and teamwork to solve real business problems under tight deadlines. Graduates emerge highly confident and bonded to each other and the company.
TU achieves more than typical corporate boot camps by serving as Trilogy's primary research and development engine and incubator for strategic thinking and next generation leadership. This is possible because Trilogy's top leadership is deeply engaged in TU, participating directly and using it to develop the company's future leaders. The program has become a virtuous cycle that continuously improves due to this leadership involvement and commitment to innovation.
The Extreme BlueTM internship program at IBM places top students on project teams to develop new products and services. Interns submit many patent disclosures and open source contributions. They help create solutions for clients and bring new IBM products to market. Unlike other internships, interns work on leading technologies like cloud computing and petabyte-scale data analysis.
The AT&T MATREX IT Internship offers interns experience in the inner workings of a technology company through challenging assignments, exposure to technology teams, and hands-on experience.
Cisco offers flexible internships globally in various business departments, from 3-18 months. Internships provide opportunities in fields like sales, engineering, finance, and marketing.
Sabine Douglas is a software and mobile development consultant who has extensive experience recruiting for these fields. She focuses on recruiting candidates with specific technical skills rather than targeting particular industries. Sabine works with organizations of all sizes, from start-ups to large companies. She has a strong presence in the North West of England but finds mobile development candidates across the UK. Recommendations praise Sabine for her specialized expertise, high-quality placements, and personal approach to understanding clients' and candidates' needs.
Presented by Neil Perlin
Considering converting your help authoring tool (HAT) output to mobile but not sure what you’re getting into? Recent releases of HATs like Flare and RoboHelp can output to multiple channels such as ebooks, web apps, HTML5, even native apps. Mechanically, it’s surprisingly simple. It’s in the interface design and information design that things can get messy. Come to this session to learn about how. We’ll cover:
The types of mobile supported by HATs and how to define your mobile needs
Interface differences between online help and mobile
What help authoring tool features work, may work, and won’t work in mobile outputs
The document provides details about the agenda for a presentation including sessions, speakers, and topics. Some of the session topics include responsive design, leadership for introverts, scenario-based design, blogging, multi-channel content publishing, volunteering to advance your career, making products interesting, the future of knowledge transfer, using cloud technology, the role of technical writers, managing client expectations, using plain language, HTML5, CSS, mobile outputs, and responsive design. The closing session will include a lightning talk on happiness and well-being and a session on technical writing for FDA-regulated industries.
The document describes several Watson services available on IBM's Watson Developer Cloud platform:
1) User Modeling extracts cognitive and social characteristics from user communications to help understand user preferences.
2) Question and Answer interprets questions and returns responses directly from source documents.
3) Relationship Extraction identifies entities and relationships within unstructured text.
This document discusses how an agile transformation can be self-funding through an incremental, evolutionary approach. It advocates bootstrapping agile practices internally by taking iterative approaches to implementing processes. This allows benefits to be realized early on, which can then be reinvested to further the transformation. It provides an example of a company that transitioned to agile in this way, initially implementing practices like Scrum and XP on their own and seeing improvements that enabled continued training investments over time.
Scenario based design 2014 mid atlantic conference Donn DeBoardddeboard
As technical communicators, we develop feature-based content that provides details about our products. But, often, this content doesn't address our customers job context to help them perform tasks successfully. Our customers search to find the right content to match their job context.
Scenario-based design helps you define a customer's work or job context using your product. This session provides of overview of scenarios and how to create them.
This document discusses how remote work is becoming more common and outlines 3 steps companies can take to create a more engaged remote workforce: 1) identify the right remote-enabling technologies; 2) establish a remote working policy; and 3) integrate remote technologies into key business functions like HR, sales, and marketing. It provides examples of how cloud technology, video conferencing, mobile devices, and enterprise social media can help attract and engage talent, streamline sales processes, and allow marketing teams to collaborate across distances.
DevOps by Design -- Practical Guide to Effectively Ushering DevOps into Any O...Dana Gardner
Transcript of a Briefings Direct discussion on some powerful best practices on making DevOps an accelerant to broader business goals, but at the level of a multigenerational IT activity.
Unum Group Architect Charts a DevOps Course to a Hybrid Cloud FutureDana Gardner
Transcript of a BriefingsDirect podcast on how Unum Group has benefitted from a better process around application development and deployment using HP tools.
Ultimate guide to hire dedicated developer comparison, benefits, & tipsKaty Slemon
Know why and when should you hire dedicated developer, learn about the essential things to consider, benefits, compare with freelancers, and tips for hiring.
This document discusses various perspectives on DevOps. It begins by noting that DevOps adoption is increasing in the enterprise and that incentivization of employees and certification are important topics. It then explores debates around whether DevOps is elitist, the challenges in recruiting for DevOps roles, the relationship between DevOps and sysadmin work, the need for DevOps training programs, and how to effectively communicate with DevOps professionals during recruitment without treating them as "unicorns, ninjas and rock stars."
This document provides an overview of considerations for implementing DevOps transformations within large enterprises. It discusses that people, processes, and technology must be aligned for a successful transformation. Regarding people, it notes the importance of organizational design, skills, and incentives/culture to attract and retain DevOps talent. For processes, it recommends evaluating business processes to ensure knowledge and responsibilities are properly allocated. Finally, it stresses that culture change is difficult but necessary, and processes should empower collaboration instead of siloed work.
This document discusses DevOps and the use of metrics and data in DevOps. It provides perspectives from DevOps leaders on some of the challenges around collecting and using large amounts of data. While having more data is possible today due to technologies like Hadoop, simply collecting more data does not necessarily provide more value and can create "noise". Useful metrics should either inform teams about issues that need to be addressed or validate hypotheses. The document discusses determining which specific metrics will provide valuable insights and avoiding collecting data just for the sake of collecting it.
This document provides an overview of considerations for implementing DevOps transformations within large enterprises. It discusses that people, processes, and technology must be aligned for a successful transformation. Regarding people, it notes the importance of organizational design, skills, and incentives/culture to attract and retain DevOps talent. For processes, it recommends evaluating business processes to ensure knowledge and responsibilities are properly allocated. Finally, it stresses that culture change is difficult but necessary, and processes should empower collaboration instead of siloed work.
DevOps Engineer Hiring Guide
Employing the ideal individuals and building a fruitful group is no simple assignment. There are
numerous aspects to consider when conversing with applicants, from social fit and group
elements to abilities, information, and critical thinking capacity. This employment guide will
address that large number of regions and that's only the tip of the iceberg. All the more critically,
this aide will assist you with exploring the one-of-a-kind elements that include DevOps
development.
If you are looking for Hire the best developer, Just Look here: DevOps Engineer Hiring Guide
Employing for DevOps ability presents its own difficulties. To begin with, your selecting group or
office has to know what to search for in applicants - there are significant attributes that may not
show up as catchphrases on a resume or in a LinkedIn profile. Second, the most qualified
applicants likely arent effectively looking for the positions employing supervisors are hoping to fill.
What's more that is the place where this aide can assist you with an exhortation from our
supporters.
DevOps Engineer Hiring Guide Employing the ideal individuals and building a fruitful group is no simple assignment. There are numerous aspects to consider when conversing with applicants, from social fit and group elements to abilities, information, and critical thinking capacity. This employing guide will address that large number of regions and that's only the tip of the iceberg. All the more critically, this aide will assist you with exploring the one-of-a-kind elements that include DevOps development. Employing for DevOps ability presents its own difficulties.
To begin with, your selecting group or office has to know what to search for in applicants - there are significant attributes that may not show up as catchphrases on a resume or in a LinkedIn profile. Second, the most qualified applicants likely arent effectively looking for the positions employing supervisors are hoping to fill. What's more that is the place where this aide can assist you with an exhortation from our supporters. Assuming that DevOps is about individuals, cycles, and instruments, then, at that point, getting the ideal individuals in your DevOps group can represent the deciding moment for the achievement of your authoritative objectives and goals.
The document discusses trends in DevOps roles in London from 2014 to 2015 based on vacancy data. It finds that demand for DevOps professionals is outpacing supply, leading to significantly higher vacancy volumes and rapidly increasing salaries. DevOps roles have grown from constituting 3% of software/engineering roles in 2014 to 4.5% currently. Demand is strongest in London, where DevOps roles now make up 5.6% of the total. Software companies and digital agencies have seen the largest increases in DevOps vacancies year-over-year.
The document discusses trends in DevOps roles in London from 2014 to 2015 based on vacancy data. It finds that demand for DevOps professionals is outpacing supply, leading to significantly higher vacancy volumes and rapidly increasing salaries. DevOps roles have grown from constituting 3% of software/engineering roles in 2014 to 4.5% currently. Demand is strongest in London, where DevOps roles make up 5.6% of the total. Software companies and digital agencies have seen the largest increases in DevOps vacancies year-over-year.
IIHT’s Integrated Learning Program for engineering students is a blend of technical skills and academic knowledge put into practice via live projects, workplace & analytical skills. The program prepares students to successfully carry out tasks and effective relationships at modern workplaces. Meant for 3rd to 7th-semester students, learners successfully completing the program becomes eligible for placement assistance by IIHT. Register as an institution.
Migrating to the cloud is a big decision. It can be difficult to know where to start, who to trust, and what the best path for your company is. That's where a cloud migration consultant comes in. A consultant can help you understand all of your options and make the best decision for your company. They can also help with the actual migration process, making sure it goes as smoothly as possible.
Presented by Neil Perlin
Considering converting your help authoring tool (HAT) output to mobile but not sure what you’re getting into? Recent releases of HATs like Flare and RoboHelp can output to multiple channels such as ebooks, web apps, HTML5, even native apps. Mechanically, it’s surprisingly simple. It’s in the interface design and information design that things can get messy. Come to this session to learn about how. We’ll cover:
The types of mobile supported by HATs and how to define your mobile needs
Interface differences between online help and mobile
What help authoring tool features work, may work, and won’t work in mobile outputs
The document provides details about the agenda for a presentation including sessions, speakers, and topics. Some of the session topics include responsive design, leadership for introverts, scenario-based design, blogging, multi-channel content publishing, volunteering to advance your career, making products interesting, the future of knowledge transfer, using cloud technology, the role of technical writers, managing client expectations, using plain language, HTML5, CSS, mobile outputs, and responsive design. The closing session will include a lightning talk on happiness and well-being and a session on technical writing for FDA-regulated industries.
The document describes several Watson services available on IBM's Watson Developer Cloud platform:
1) User Modeling extracts cognitive and social characteristics from user communications to help understand user preferences.
2) Question and Answer interprets questions and returns responses directly from source documents.
3) Relationship Extraction identifies entities and relationships within unstructured text.
This document discusses how an agile transformation can be self-funding through an incremental, evolutionary approach. It advocates bootstrapping agile practices internally by taking iterative approaches to implementing processes. This allows benefits to be realized early on, which can then be reinvested to further the transformation. It provides an example of a company that transitioned to agile in this way, initially implementing practices like Scrum and XP on their own and seeing improvements that enabled continued training investments over time.
Scenario based design 2014 mid atlantic conference Donn DeBoardddeboard
As technical communicators, we develop feature-based content that provides details about our products. But, often, this content doesn't address our customers job context to help them perform tasks successfully. Our customers search to find the right content to match their job context.
Scenario-based design helps you define a customer's work or job context using your product. This session provides of overview of scenarios and how to create them.
This document discusses how remote work is becoming more common and outlines 3 steps companies can take to create a more engaged remote workforce: 1) identify the right remote-enabling technologies; 2) establish a remote working policy; and 3) integrate remote technologies into key business functions like HR, sales, and marketing. It provides examples of how cloud technology, video conferencing, mobile devices, and enterprise social media can help attract and engage talent, streamline sales processes, and allow marketing teams to collaborate across distances.
DevOps by Design -- Practical Guide to Effectively Ushering DevOps into Any O...Dana Gardner
Transcript of a Briefings Direct discussion on some powerful best practices on making DevOps an accelerant to broader business goals, but at the level of a multigenerational IT activity.
Unum Group Architect Charts a DevOps Course to a Hybrid Cloud FutureDana Gardner
Transcript of a BriefingsDirect podcast on how Unum Group has benefitted from a better process around application development and deployment using HP tools.
Ultimate guide to hire dedicated developer comparison, benefits, & tipsKaty Slemon
Know why and when should you hire dedicated developer, learn about the essential things to consider, benefits, compare with freelancers, and tips for hiring.
This document discusses various perspectives on DevOps. It begins by noting that DevOps adoption is increasing in the enterprise and that incentivization of employees and certification are important topics. It then explores debates around whether DevOps is elitist, the challenges in recruiting for DevOps roles, the relationship between DevOps and sysadmin work, the need for DevOps training programs, and how to effectively communicate with DevOps professionals during recruitment without treating them as "unicorns, ninjas and rock stars."
This document provides an overview of considerations for implementing DevOps transformations within large enterprises. It discusses that people, processes, and technology must be aligned for a successful transformation. Regarding people, it notes the importance of organizational design, skills, and incentives/culture to attract and retain DevOps talent. For processes, it recommends evaluating business processes to ensure knowledge and responsibilities are properly allocated. Finally, it stresses that culture change is difficult but necessary, and processes should empower collaboration instead of siloed work.
This document discusses DevOps and the use of metrics and data in DevOps. It provides perspectives from DevOps leaders on some of the challenges around collecting and using large amounts of data. While having more data is possible today due to technologies like Hadoop, simply collecting more data does not necessarily provide more value and can create "noise". Useful metrics should either inform teams about issues that need to be addressed or validate hypotheses. The document discusses determining which specific metrics will provide valuable insights and avoiding collecting data just for the sake of collecting it.
This document provides an overview of considerations for implementing DevOps transformations within large enterprises. It discusses that people, processes, and technology must be aligned for a successful transformation. Regarding people, it notes the importance of organizational design, skills, and incentives/culture to attract and retain DevOps talent. For processes, it recommends evaluating business processes to ensure knowledge and responsibilities are properly allocated. Finally, it stresses that culture change is difficult but necessary, and processes should empower collaboration instead of siloed work.
DevOps Engineer Hiring Guide
Employing the ideal individuals and building a fruitful group is no simple assignment. There are
numerous aspects to consider when conversing with applicants, from social fit and group
elements to abilities, information, and critical thinking capacity. This employment guide will
address that large number of regions and that's only the tip of the iceberg. All the more critically,
this aide will assist you with exploring the one-of-a-kind elements that include DevOps
development.
If you are looking for Hire the best developer, Just Look here: DevOps Engineer Hiring Guide
Employing for DevOps ability presents its own difficulties. To begin with, your selecting group or
office has to know what to search for in applicants - there are significant attributes that may not
show up as catchphrases on a resume or in a LinkedIn profile. Second, the most qualified
applicants likely arent effectively looking for the positions employing supervisors are hoping to fill.
What's more that is the place where this aide can assist you with an exhortation from our
supporters.
DevOps Engineer Hiring Guide Employing the ideal individuals and building a fruitful group is no simple assignment. There are numerous aspects to consider when conversing with applicants, from social fit and group elements to abilities, information, and critical thinking capacity. This employing guide will address that large number of regions and that's only the tip of the iceberg. All the more critically, this aide will assist you with exploring the one-of-a-kind elements that include DevOps development. Employing for DevOps ability presents its own difficulties.
To begin with, your selecting group or office has to know what to search for in applicants - there are significant attributes that may not show up as catchphrases on a resume or in a LinkedIn profile. Second, the most qualified applicants likely arent effectively looking for the positions employing supervisors are hoping to fill. What's more that is the place where this aide can assist you with an exhortation from our supporters. Assuming that DevOps is about individuals, cycles, and instruments, then, at that point, getting the ideal individuals in your DevOps group can represent the deciding moment for the achievement of your authoritative objectives and goals.
The document discusses trends in DevOps roles in London from 2014 to 2015 based on vacancy data. It finds that demand for DevOps professionals is outpacing supply, leading to significantly higher vacancy volumes and rapidly increasing salaries. DevOps roles have grown from constituting 3% of software/engineering roles in 2014 to 4.5% currently. Demand is strongest in London, where DevOps roles now make up 5.6% of the total. Software companies and digital agencies have seen the largest increases in DevOps vacancies year-over-year.
The document discusses trends in DevOps roles in London from 2014 to 2015 based on vacancy data. It finds that demand for DevOps professionals is outpacing supply, leading to significantly higher vacancy volumes and rapidly increasing salaries. DevOps roles have grown from constituting 3% of software/engineering roles in 2014 to 4.5% currently. Demand is strongest in London, where DevOps roles make up 5.6% of the total. Software companies and digital agencies have seen the largest increases in DevOps vacancies year-over-year.
IIHT’s Integrated Learning Program for engineering students is a blend of technical skills and academic knowledge put into practice via live projects, workplace & analytical skills. The program prepares students to successfully carry out tasks and effective relationships at modern workplaces. Meant for 3rd to 7th-semester students, learners successfully completing the program becomes eligible for placement assistance by IIHT. Register as an institution.
Migrating to the cloud is a big decision. It can be difficult to know where to start, who to trust, and what the best path for your company is. That's where a cloud migration consultant comes in. A consultant can help you understand all of your options and make the best decision for your company. They can also help with the actual migration process, making sure it goes as smoothly as possible.
10 Futuristic Leaders Mastering the DevOps Universe, 2023.pdfCIO Look Magazine
This edition features a handful of The Futuristic Leaders Mastering the DevOps Universe that are leading us into a better future
Read More: https://ciolook.com/10-futuristic-leaders-mastering-the-devops-universe-2023-july2023/
Accelerating Employee Onboarding & Your Hybrid Workplace With Microsoft 365Richard Harbridge
Join us and learn how you can deepen your existing technology investment to support employee onboarding and your hybrid workplace.
A whopping 73% of employees want flexible remote work options to stay, says Microsoft in their latest report, The Next Great Disruption is Hybrid Work. This makes tailoring your employee onboarding to support the hybrid working model essential to the success of your team!
As such, many organizations are making the adjustments for a post-pandemic world and looking to prepare for and manage the explosive growth expected in the months and years ahead. But, from skills development and digital excellence to process automation and connecting employees to the resources they need at the right time, the challenges ahead are considerable.
Join LiveTiles along with LiveTiles Partner, 2toLead, CTO and Microsoft MVP, Richard Harbridge on Tuesday, May 11, from 11:00 AM to 12:00 PM EST as he explores:
How the “workplace” has changed
Why digital excellence should be prioritized in onboarding, especially when supporting hybrid work and the remote workforce
Ways organizations are looking to better leverage Microsoft 365 and their digital workplace investments to support this talent growth
How SharePoint, Microsoft Teams, Yammer, and more can be used with support from LiveTiles solutions (pictured below) to improve the employee onboarding and hybrid work experience
2i recently attended a DevOps Summit in London to learn more about how different companies have implemented DevOps. Read our overview to gain a better understanding of the DevOps operating model.
How to hire a DevOps engineer, you might look to your IT Operations groups for ability. Normally, these groups have focused on critical thinking. Because of the expansion in intricacy and extent of activities, individuals from these groups have likely needed to shuffle a lot of work with individuals from different foundations. These people make ideal DevOps designers of things to come because of their much of the time assorted foundations and knowledge. Also, advanced architects can be great up-and-comers since they comprehend designing standards and have innovation proficiency. Discover how they can work together to make new items or administrations and how they can share their inventive attitude.
DevOps is an exciting new management framework that combines software development and IT operations. It aims to shorten the systems development life cycle and provide continuous delivery with high software quality. DevOps is rapidly popularity across the IT industry due to the ease with which it can be used in combination with Agile software development.
Original Source: https://www.knowledgetrain.co.uk/it/devops/what-is-devops
Patterns for Success: Lessons Learned When Adopting Enterprise DevOpsCognizant
The document discusses common reasons for failure of DevOps initiatives in large enterprises and provides recommendations for successful adoption of DevOps. Some key reasons for failure include lack of a common definition, organizational resistance to change, cultural issues, technology complexity, divergent tools used, architectural differences, and existing technical debt. The document recommends addressing these issues by having a well-defined plan, executive support, stakeholder buy-in, dedicated roles to lead the effort, a phased approach using pilots, automation, and objective metrics to track progress. Following these patterns can help large organizations successfully adopt DevOps.
Building a strong digital culture can feel like an impossible endeavor. How do we create virtual workplaces with a shared sense of values? How do we sustain mentorship relationships and strengthen information sharing protocols? Without a good framework in place, it’s extremely unlikely that a team can build a hybrid culture, let alone a completely remote workplace. To leverage remote work, every organization must employ tried-and-true precepts.
Convergence - Diverse Journeys to the Same Truthjack_maher
We are all pilgrims on a common journey with many shared paths on our way to improving our capabilities and helping our organizations create and deliver value.
Presented and discussed at Agile Cincinnati on June 11, 2020.
Importance of Building a DevOps Culture for Successful Digital Transformation...Urolime Technologies
In today's rapidly changing digital landscape, businesses must adapt and transform to remain competitive. However, digital transformation is not just about adopting new technologies, but also about creating a culture that promotes collaboration, agility, and innovation. This is where DevOps comes in - a methodology that emphasizes cross-functional teams, continuous integration and delivery, and automation. By establishing a DevOps culture, organizations can improve their speed, quality, and reliability of software delivery, ultimately driving business success. In this article, we will explore the key benefits and elements of implementing a DevOps culture in your digital transformation strategy, and how it can help your organization stay ahead of the curve.
2. CONTENTS
Introduction 3
Is DevOps too elitist? 4
DevOps – the recruitment challenge 8
DevOps and the bottom line 11
Microservices and data consistency 14
DevOps institutional thinking 16
What is ‘micro’ about a microservice? 20
Flipping incentivization 23
Educating DevOps 26
Wielding the double-edged sword of automation 29
devops perspectives 3
3. 3
There’s some evidence that DevOps as a core part of
mainstream IT strategies is not that far away, although as
Em Campbell-Pretty points out on page ??, there’s still a lot
of ‘de-training’ to be done with the CIO first!
Recent articles predicting the increased enterprise adoption
of DevOps and the formation of the DevOps Institute
point to a maturation of DevOps as an approach. One key
for adoption is incentivization and how employees are
objectivized on success and how this affects their behaviors,
which James Smith of DevOps Guys explores on page x.
Equally, the creation of the DevOps Institute has created
some stir in the market, with views for and against the
need for some form of certification of capabilities in this
area being expressed widely. Read more about this debate
on page x. Microservices as a hot topic has come to the
forefront in the last 12 months or so and it’s one we have
not just one but two expert opinions on, with Matthew
Skelton and Jason Bloomberg providing the consultant and
analyst views on pages x and x respectively.
There’s some evidence that DevOps as a core part of mainstream IT strategies is not that
far away, although as Em Campbell-Pretty points out on page ??, there’s still a lot of ‘de-
training’ to be done with the CIO first!
devops perspectives 3 | introduction
DevOps - mainstream
tipping point ahead
One popular definition of DevOps uses the acronym
“CALMS”; standing for the topics of Collaboration,
Automation, Lean, Measurement and Sharing. The
“Automation” subject is one which has seen a great level of
buy-in with various vendors offering automated capabilities
in the build, continuous integration, environment
provisioning and release management areas for example
(CA Technologies being one of these of course). Dan North
explores where and how automation is a good thing, and
also advises where to adopt a more cautionary approach
in his excellent article on page x. Lastly, let’s not forget
that the DevOps philosophy hinges on people, process,
technology and information, an argument also made by
Nicole Forsgren on page x.
We hope you find this issue as stimulating and thought-
provoking as previous editions.
4. 4
There’s a bone of contention in the industry
that those in the DevOps space too often
display what we might call tech elitism. I’d
counter that and suggest that what we’re
actually talking about is just the latest group
of talented IT folk in an exploding market,
who are riding the current tech and cultural
tsunami to their benefit.
Is DevOps
too elitist?Paul Speers, CEO, Speerhead
In fact, I’d argue that what appears to be happening in the
DevOps market is no different to the SAP market in the late
80s or the Siebel CRM wave in the 90s.
The success of those vendors spawned the creation of
communities of technical wizards and magicians. It also
created a lot of wealth for certain people—and earning twice,
even three times, the average salary in your field inevitably
breeds a certain type of superiority.
In a sense, DevOps is going through the same paradigm shift,
but this is about people, smarts and talented skills at the
forefront of the market rather than about a vendor application.
devops perspectives 3 | Is DevOps too elitist?
5. 5
One of the biggest hiring challenges in the DevOps
market right now is around job descriptions. Most of
those that we see are shockingly light on detail and
quite simply lack content. As a result, HR folk are lost
at sea, with much vague arm-waving in the air when
attempting to describe the requirements of the role.
The consequence of this is that hiring managers in the
main are engineers, who are able to articulate what they
need, and are trying to achieve, and to the right sort of
person who can understand what they are talking about.
But very rarely is this expertise tangibly reflected in a
briefing document or even a briefing session. Mostly it’s
a woolly four-paragraph job spec for a £95,000 package.
Is it surprising then that candidates with these skills
are so sought after and constantly hounded, because
hardly any recruiter or HR department really knows how
to speak DevOps or understand what the businesses are
trying to achieve?
Part of the rise of DevOps demand stems from the
notion that no one is hiring sysadmins any more, it’s
all about DevOps. Over the past 10 years, sysadmin has
become a dirty word.
As market dynamics kick in and the people grab
acronyms, sysadmin is not an alluring job title anymore.
On the other hand, DevOps is very ‘spicy’ so that’s what
people call themselves. But on closer inspection, in
many cases the job descriptions are the same.
The same thing happened to me when the clamour
for LAN and WAN networking skills happened. I went
from being a £250-a-day systems manager contractor
to earning £1200 a day as a sysadmin in the space of
a week, even though I wasn’t trained in sysadmin and
operational skills.
So, on the basis that salaries are rising, why not just
change your job title to the current market favorite
because it’s the type of work that’s in demand? Who
wouldn’t do this, given that it’s market growth and our
industry’s appetite for acronyms that’s driving it?
The end result is that DevOps is not a lower-level job
title anymore. The good candidates have transitioned
into engineering roles, designing complex cloud
systems, bashing out intricate scripts, engineering clouds
and creating elastic automated infrastructures. These
are good and clever people who contribute value to the
organization.
But that’s not uniformly the case. Inevitably there are
also the ‘rock stars and ninjas’ to be dealt with.
What’s in a name? Same stuff, different time
devops perspectives 3 | Is DevOps too elitist?
“Part of the rise of DevOps demand stems from the notion that
no one is hiring sysadmins any more, it’s all about DevOps”
6. 6
The ‘rock star’ definition in this context means you will
be one person doing the job of an eight-person team, but
is this a hero culture? Maybe in some businesslike start-
ups, it’s totally relevant. However, you might prefer to hire
session musicians or even lead tenors in a choir than have
an Axel Rose-style hellraiser in the team.
We have seen a couple of distinct categories emerge
when we’ve been asked to headhunt people. To hire for a
4–8 person start-up, you need a hero or rock star, living for
something to fail, constantly bootstrapping and working
24 hours a day. But in an enterprise team environment,
you don’t need heroes, you need different calibres of
individuals. But where to find them?
As the DevOps contractor market has boomed, the talented
folk are migrating from full-time to contract status, thus
creating a vacuum in the middle of the market with the
lack of skills.
That lack of skills means that enterprises are overpaying for
those that are in play. This sort of situation can contribute
to a certain inflated opinion of your own value.
So what’s the answer? We need to see DevOps training
programs at scale on new tech, hard and soft skill sets.
Training programs are required for both the new grads
who are starting out, and also for seasoned domain-skilled
engineers who want to up-sell and cross-sell their careers.
But there are some rays of light when it comes to the move
to the DevOps title. There are limited resources available to
train for this type of DevOps job, but Ops School has started
the journey.
In addition, the explosion of internet businesses has
triggered a mind shift, with start-ups needing people with
an operational mindset, who can not only develop and cut
code, but keep all the lights on and sustain the company.
Ninjas and rock stars
NOT required
Skills crisis
“You don’t need heroes,
you need different
calibres of individuals.”
devops perspectives 3 | Is DevOps too elitist?
7. 7
Not everyone gets it right. Our candidates send us some very
funny emails from ‘old school’ recruiters who know nothing
about DevOps! This low-cost, easy-to-use communication
method just adds to the awareness of the entire DevOps
resource community that they are in demand.
Now add a large dose of IT ‘reccers’ where it is all about the
numbers and the ‘old school’ recruitment KPIs. These mass
emails and irrelevant jobs derive from outdated techniques
and give the wrong impression. Also no passive candidate in
their right mind will respond to this as it’s just wasting their
time when the conversation is non-DevOps-savvy.
So how do we all calm things down a bit? Sadly I don’t think
we can, but when we start working on a role for a business, our
messaging is very particular.
Each candidate gets a special communication ‘snowflake’ sent
to them, that is highly relevant and interesting, calling out skills
in their past that are key assets for the next role.
This enables us to start a proper dialogue and connection
about the job in detail. We read the public domain profile and
understand where they have come from.
We have built our business on knowing our audience—and yes,
DevOps architects and engineers are very, very special folk—
and engaging in a proper relationship understanding skills,
tools and career advancement.
Overall, what we see is a need for a new mantra that is NOT
about hiring unicorns, ninjas and rockstars. All we need are
good solid developers with cloud tech experience who can
relate to the business goals and customer experience.
Those people are clever folk who frankly have a right to treated
in a different way to others as they are leading change and
driving business agility. If that’s elitist, so be it.
Now bother DevOps folk
with spamming… right?
How to calm
the madness down
“it’s NOT about hiring
unicorns, ninjas and
rockstars”
devops perspectives 3 | Is DevOps too elitist?
8. 8
Grant Smith, author of Next Gen DevOps, can understand
how the elitist charge against DevOps has come about.
“When a less powerful group challenges the perceptions
and privileges of a more powerful group, there are always
accusations and misinterpretation,” he says.
“DevOps demands that development and test teams
become involved in the management of services
throughout their lifespan. Those CTOs coming to their role
from a product development or management background
struggle with this concept because they don’t understand
the difference between an application and a service.”
Tony Chapman, Managing Director at Linux Recruit, poses
an intriguing question when he asks: “Is being elitist
necessarily always a negative thing?
“Any new movement at the cutting edge needs leaders to
define the approach and disciples to follow. DevOps needs
the Elite to master the approach and teach others; as long
as those at the pinnacle allow others to follow, it should be
healthy for the DevOps community.”
Patrick Hyland, Founder of DevOps Associates, talks less
about elitism and more about dedication above and
In the previous article, Paul Speers, CEO of Speerhead, addressed the
charge that DevOps is in danger of behaving in an elitist manner
and argued that good DevOps people are entitled to be treated
differently. But is this a view that finds favor across industry?
And is the ‘elitist’ argument masking a more concerning issue—that
of organizations pursuing the elite DevOps title at the expense of
missing out on highly experienced candidates who are rejected due
to not having the right job title?
DevOps –
the recruitment challengeGrant Smith, Author, Next Gen DevOps
Tony Chapman, Managing Director, Linux Recruit
Patrick Hyland, Founder, DevOps Associates
“Is being elitist
necessarily always a
negative thing?”
devops perspectives 3 | DevOps – the recruitment challenge
9. 9
beyond the norm. “To be successful in any discipline, to go
beyond mere mediocrity, one needs to be committed to,
passionate and very serious about the work,” he argues.
“The Myers Briggs MBTI personality theory may be useful in
understanding some of allegations of elitism, both from the
point of view of an individual who interprets an interaction
as elitist and from the point of view of the person who is
being viewed as elitist,” he adds.
“There are obvious, maybe objective cases where a person
comes across as arrogant and condescending but I think
outside of those obvious cases there is a much more
nuanced personality-specific interpretation happening,”
he concludes. “The best of these individuals are tolerant,
have developed a mature self-awareness and know how to
approach the work with humility and a sense of balance.”
They’re also well paid. Chapman sees the higher salaries
on offer to those with DevOps job titles as a simple case
of market economic realities. “The daily rate debate within
any technical industry is a simple case of supply and
demand,” he argues. “As with anything, if demand exceeds
supply, prices rise.
“There is a huge demand for an intricate set of skills
within the DevOps toolset, and a short supply of highly
qualified individuals. This invariably increases rates. Due
to high rates, the contract DevOps market is beginning
to be flooded with contractors adding ‘DevOps’ into their
job titles, or other buzzwords in technologies: Docker,
Puppet, Ansible etc.”
This is hardly surprising, he adds: “If someone is willing
to pay people an inflated rate, purely by adding DevOps
into their title, you can understand it to a certain degree.
The issue is uneducated recruitment consultants, or even
uneducated hiring organizations, not understanding
what is required to be a true DevOps individual.”
Mature DevOps practitioners are rare but the demand
is high, agrees Hyland, adding that there’s also a good
“As with anything, if
demand exceeds supply,
prices rise.”
devops perspectives 3 | DevOps – the recruitment challenge
10. 10
deal of confusion around what DevOps really means in
practice when it comes to recruiting staff.
“There is still too much focus put on technology skills and
too little put on mature system thinkers with a technical
background,” he says. “I think if clients and recruiters
phrased the job spec in a different way they may get more
of the types of candidates that they and the clients’ people
want to work with.”
This is a point picked up by Smith, who makes the case
that DevOps is not a monoculture and, given that people
move into the space from different places, there’s not an
accepted single perception of what DevOps actually is.
“The IT industry as a whole is confused about what DevOps
is and what the advantages are and that goes double for
CTOs, most of whom started their careers in development
or product,” he says. “DevOps is challenging the status quo.
It is focusing people’s attention on the service as a whole,
not just on application development and launch.”
Smith flags up shortcomings in how recruiters source
DevOps candidates. “Recruiters have optimized their
sourcing processes so that their tools or their low-paid
sourcers can scan through thousands of CVs and LinkedIn
profiles, pattern matching for computer science, Java, Git,
Jenkins etc,” he explains.
“The recruiter can then contact the developer with an
almost cookie-cutter job description and the developer
will understand what they’ll be expected to do. There is no
DevOps or Operations degree. There aren’t standardized
languages or tools yet and so recruiters have a harder job
identifying candidates. That means that recruiters are
bombarding anyone with DevOps on their CV for every role
they have.”
But there’s a lot of DevOps opportunism out there that
can pose a problem, argues Chapman. “Being elitist as an
organization, without actually understanding what this
means, can potentially render some positions literally
unfillable,” he warns.
“There are companies we have partnered with, who are
not as educated or advanced enough in the DevOps arena
but who engage us requiring a DevOps engineer. On
further qualification, it’s clear they just require a system
administrator (or another skill set).
“But the system administrators are discounted due to not
having ‘DevOps’ written on their CV. The actual DevOps
guys are not interested due to on further inspection it not
actually being a true DevOps role. It continues in a cycle,
until system administrators need to include ‘DevOps’ on
their CV to be considered, which they are then potentially
vindicated for.”
This could have long-term negative implications that stem
from the pursuit of this elitist status. “There is the possibility
of some organizations missing out on great people, purely
due to them not having the latest buzz technology listed as
a skill,” suggests Chapman.
“I see a huge amount of organizations rejecting very good
candidates for a permanent position because, for instance,
they don’t have Puppet experience. For a talented engineer,
learning a new technology such as Puppet configuration
management shouldn’t be the hardest thing in the world,
as long as they have a solid engineering background and
the right mindset.
“Many of the talented Linux sysadmins I have previously
worked with and placed into great organizations are now
among the elite DevOps Engineers and visionaries. At
some point in the last few years they have learnt how to
master CI/CD, Automated Deployments and Configuration
Management tools, so arguably we need to give others this
opportunity without immediately discounting them.”
He concludes:
“Many companies are creating dangerous situations where
they are rejecting potentially great people, who could
provide a significant long-term ROI, simply due to them not
having the latest buzzword on their CV.”
“Being elitist can
potentially render
some positions literally
unfillable”
devops perspectives 3 | DevOps – the recruitment challenge
11. 11
This, she argues, flies in the face of decades of research.
“For the first time in recent times, research shows a link
between IT investment and organizational performance,
but only if those investments exist with the right mix of IT,
culture and practice, aka DevOps,” she says.
“What’s interesting is that for decades we’ve tried to find
the value that technology can bring to a business and
haven’t been able to find it. Investment in IT doesn’t
impact the bottom line, any kind at all. We just don’t see it.
Studies fail to show a link, time and time again.”
This is a manifestation of what Forsgren calls the
Productivity Paradox. “Anyone can go out and buy a
server and throw it in the closet,” she explains. “But your
competitors can do that as well. It’s a low barrier to entry,
so they can buy the same server or a similar server and so
you don’t get any real competitive differentiator.
“If you do get any kind of competitive advantage, it’s
just not sustainable because technology advances. So
technology never becomes a differentiator for you. What
you need to do is find a way to leverage technology—or
indeed anything you have—in such a way that you can
really set yourself apart from your competitors. The ROI
rarely pans out and if it does, you’re looking at three, four or
five years. It just doesn’t work.”
But DevOps is different, suggests Forsgren, stating that:
“DevOps is not just a technology solution. It’s not just a
server, it’s a major re-engineering shift. DevOps has ended
up being a significant process change. Yes, it requires IT
investment, but it also requires investment in culture and
the re-engineering of processes.
“For change to happen, you have to include the right people
and processes, the right culture, the right tools, the right
technology. You need to have a culture in place that is open
and generative and communicative. You need to have Dev
and Ops talking. You have to re-engineer the entire process
and you have to do it very mindfully. We don’t talk about
best practice, but you need to have good practices.
“It’s like the lean and the Toyota way of manufacturing that
we saw in the 1990s. Just as that revolutionized the way
manufacturing was done, DevOps will revolutionize the way
IT is done across all industries.”
DevOps is not just about making an impact on the IT function, but also on the bottom line—
and that’s revolutionary. That’s the hypothesis proposed by Nicole Forsgren PhD, Assistant
Professor at Utah State University and Director of Organizational Performance and Analytics
at Chef, and she’s got the data to back it up.
DevOps and the bottom line
Nicole Forsgren, Director of Organizational Performance and Analytics, Chef
“You need to have a
culture in place that is
open and generative
and communicative.”
devops perspectives 3 | DevOps and the bottom line
12. 12
What’s also remarkable about DevOps is that its potential
reaches across both the ‘unicorn’ and ‘horse’ categories
of company. “Unicorns are young and nimble. They are
start-ups or they were start-ups. They’re companies like
NetFlix and the way that they do things is the DevOps way,”
explains Forsgren.
“Horses are major established companies which have
been around for so long that they are set in their ways. It’s
like old dogs and new tricks. You have firms that just can’t
drastically change the way they do things.”
Falling under the ‘unicorn’ banner would be a firm like
accounting software provider Intuit, which used DevOps to
experiment with new functionality. What was particularly
bold was that this experimentation took place during the
tax season, the busiest period for the company.
Again, this runs counter to accepted wisdom, but in reality
what time is better to do experiments with functionality
than the period during which customers are using your
products most? In this case, following such a course of
action had a business impact of delivering a 50 percent
increase in conversion rates on the website.
Meanwhile at a ‘horse’ company, DevOps has resulted in
the ability to deploy code quicker and deliver services to
their customers much faster. “They can run experiments
in real time so that they can understand what features
are more valuable and decrease customer churn,” says
Forsgren. “They’re deploying code hundreds or thousands of
times a year rather than a couple of times.”
Riding the unicorns and horses
devops perspectives 3 | DevOps and the bottom line
13. 13
DevOps isn’t just IT, it’s the practice of IT. High-performing
IT organizations are twice as likely to exceed the
business’s profit, productivity and market share goals,
notes Forsgren. “You see a change in the business view of
IT. It’s seen as a cost center at first, where you have to do
IT just to keep up, but then it starts to be seen as a point
of distinction that can deliver genuine value to customers.
You can attract new customers and retain existing ones.”
Forsgren observes that DevOps has really been a ground-
up, rather than centrally driven, movement. What is now
known as DevOps has existed inside companies that
have been the best IT performing companies, but just
not called DevOps.
“Once it had a name, it became a thing,” says Forsgren. “But
it wasn’t something that came approved from the front
page of the Harvard Business Review. It came about from
people going to conferences and talking to one another
about what they’re doing. It’s one of the interesting things
about the movement: those involved in DevOps talk to each
other, and they help each other. They call one another up on
the phone and chat, or they go to meet-ups. They leverage
one another. They make it happen and then they go public
with what they’ve been able to accomplish.
More than just IT
“DevOps creeps into enterprises. You might not be able
to use DevOps across the whole organization all at once,
but it gets rolled out a piece at a time. Let’s try it here,
let’s try it there, strategically. You might identify certain
applications that are legacy and are always going to
say legacy from a business point of view. But there
are other bits of the business that can become points
of distinction, so those can use DevOps practices and
principles to deliver value to the business.”
But DevOps is good for the IT function. According to data,
high-performing DevOps teams are more agile, with reports
of 30x increase in deployments and 8000x faster lead
times than peers. They are also more reliable with 2x the
change success rate and 12x faster mean time to recovery.
So are CIOs already highly excited by the promise of DevOps
then? Not so much, according to Forsgren. “CIOs still need to
be educated when it comes to DevOps,” she suggests. “A lot
of CIOs are still running IT as simply cost centers. They just
aren’t excited by DevOps and they just don’t really care.
“It is changing though. I’ve had a few people ask me to
write up a quick blurb on evidence that it can help with
the contribution to the bottom line. Before it used to be
‘why bother, why care?’ So there is progress.”
Looking ahead, Forsgren reckons that effective use of
DevOps will become a critical business differentiator. “I can
see DevOps being part of the strategy for the enterprise,”
she says. “The smart way to do this is piece by piece. The
best way to use it is to take a strategic application and
deliver value all the way through the value chain.”
But she concludes that there will be enterprises that will
choose not to use DevOps in any way. “Some companies
will survive because they are so big,” she says. “Other
companies that resist the DevOps change just won’t
be around. Right now, adopting DevOps is a point of
distinction. For some companies it will be a point of parity.
For others, they will just fail. Survival isn’t mandatory.”
devops perspectives 3 | DevOps and the bottom line
“DevOps isn’t just IT,
it’s the practice of IT.”
14. 14
Microservices have rightly become a good pattern to adopt when deployability and rapid changes to independent services are important
for a software system. However, without a suitable data strategy in place, organisations building microservices risk data duplication and
inconsistency where coordination between service teams is limited.
Microservices and data consistencyMatthew Skelton, Co-founder and Principal Consultant at Skelton Thatcher Consulting
For software systems that need to change rapidly and often,
the microservices pattern is an emerging good practice,
particularly when combined with container technologies
such as Docker or Rocket. Not only does deployability
increase, but also Dev team engagement, together leading
to more maintainable and evolvable systems. At Skelton
Thatcher Consulting, we have been working since mid-2014
with teams at several different clients that are moving
towards smaller deployment units (whether microservices
or simply smaller, less coupled services) and we have seen
first-hand the benefits for teams. The reduction in change
complexity is a particularly useful outcome, along with a
reduction in accidental coupling.
Unfortunately, in the rush towards microservices, some
teams seem to be unaware of the value previously provided
by their older monolithic architecture, particularly if a
central relational database was used. In many organisations
with multiple product or budget streams, it is (sadly) fairly
common for the different product streams – or even multiple
‘projects’ within a single stream – to be effectively in
competition with one another.
Now, with a monolithic architecture and especially with a
central relational database, these conflicts are resolved at
compile time, data load time, or integration test time: at any
rate, almost certainly before the software reaches Production.
Feature requests that conflict (perhaps on a database column
name) are batted back to the Product Owners to resolve;
in effect, the monolith and central relational database are
acting as an arbitrator of feature requests.
However, in a microservices world, the coordinating effect of
the monolith or central relational database are typically lost,
potentially allowing changes that conflict at a fundamental
business level to reach Production before the conflicts are
detected. We have seen on several occasions that product
and programme managers can be vehemently opposed to
deliberate coordination with other product or programme
managers, preferring to pretend that their requirements are
entirely independent of those of their peers.
devops perspectives 3 | Microservices and data consistency
The hidden value of a
monolith or deployability
Optimising for
deployability
15. 15devops perspectives 3 | Microservices and data consistency
At the time of writing we have yet to see major data
consistency problems emerge from the (mis-)use of
microservices, although people like Simon Brown are
warning that the use of microservices is not a way to
avoid good software design. We are likely too early in
the adoption of microservices for the problems of data
consistency to have emerged for many teams. However,
unless different product teams are using some kind of
shared bus or event store for cross-service consistency, we
predict that the lack of a monolith or central database could
result in teams duplicating data and logical entities across
different data silos. Naturally, with a relational database, we
can rely on foreign key (FK) relationships and constraints in
order to enforce data consistency, but we have much softer
constraints in place when working with microservices and
split data storage.
This means that the organisation can no longer rely on a
central relational database for maintaining data consistency
but must find alternative methods. Perhaps the role of
the ‘data architect’ evolves to cover pan-organisational
data consistency, and maybe we’ll see some new tools
emerging over the next few years to deal with the
difficulty of correlating data from the data silos of different
microservices.
The issue of pan-organisational data consistency and
data integrity is something that cannot be ignored if
business outcomes are to be met and sustained over the
coming years, even as organisations adopt patterns such
as microservices to help achieve more rapid and frequent
software changes.
Prediction: data strategy
for microservices
traditional team of Dev, QA, DBA, Ops
Figure 1
microservices teams with a
pan-organization data strategy
Figure 3
microservices teams with data
capability within each team
Figure 2
16. 16
DevOps
institutional thinking
devops perspectives 3 | DevOps institutional thinking
The Institute’s mission statement says that it will work
with thought leaders from the DevOps and IT Service
Management communities, as well as the IT training
market to become the standard in quality, enterprise
grade DevOps education.
The DOI founders acknowledge that they are causing a
stir. On its website, the Institute concedes: “We recognise
that certified DevOps training is new to the DevOps world.
But the time has come for DevOps to take its next step.”
“To succeed best practices need to be codified and taught
in the time honoured methods used by IT for over 40
years. Not that DevOps isn’t different, it most certainly
represents a new way of thinking, doing and acting within
IT and organisations as a whole. But that doesn’t mean
traditional education and training, including certifications
will not be applicable.”
Justin Vaughan-Brown, CA Technologies
That’s the kind of bold statement that was always going
to spark heated debate in some quarters. Certainly not
everyone sees this as a good thing. Sam Newman at
ThoughtWorks, takes particular exception to the emphasis
on certification in the Institute’s statements of intent.
In a blog posting he writes: “I certainly have no problem
with people making money from DevOps in general or
DevOps training in particular. If the group had come out
and said ‘we’re going to offer awesome training courses -
look at our fantastic content!’ I would probably have been
very supportive, and may even have pointed people in
their direction. However, the Institute sees certification as
essential, and it is front and centre in their marketing.”
Heated debate
In March 2015, I hosted one of our regular “DevOps Dinners” in London and one of the
questions I posed to the assembled audience of customers, partners and thought leaders
was whether the newly launched DevOps Institute was a necessary addition to the
DevOps movement.
17. 17
Newman argues the case that DevOps is a cultural
movement, built around individuals. As such he questions
how traditional certification can measure this: “I do have
some respect for some sorts of certification in IT, but very
little of it. I see this in the same class as making sure the
person who installs my gas boiler knows their stuff.”
He questions whether you can have faith in the DevOps
abilities of someone who has a certificate of the type offered
by the Institute. Indeed he ponders whether there might an
ulterior motive at work here.
“A cynic might wonder aloud of the use of certification is
a way for trainers to set up a closed shop and make more
money as a result…The problem is that there is always going
to be a conflict of interest when the same group that sets up
the certification process controls who can deliver certified
training, at the same time as benefiting financially from
delivering training themselves.”
He concludes: “Go ahead and make money from DevOps
with my blessing (not that anyone is asking for it). Offer
training, set up whatever institute you want. But please,
don’t claim this sort of certification is going to fix the
problems we face in the Industry.”
devops perspectives 3 | DevOps institutional thinking
“DevOps is a cultural
movement, built
around individuals”
18. 18
James Smith at DevOpsGuys is similarly a naysayer. “I
fundamentally disagree with the whole concept as it
currently stands and I think the main challenge is that
there’s still no real agreement about what DevOps is,” he
says. “We have many people stating that it’s not about
tools, it’s a philosophy, a professional movement in which
case what exactly do you certify?
“If it is about ‘the cultural and professional movement that
stresses communication, collaboration, integration and
automation’ as the DOI website says, then what credentials
do the DOI have to state that their way of DevOps is the
right way? There is no one-size fits all here.”
He adds: “I also think that we are looking at the wrong
things with DOI. If DevOps is underpinned by Agile, CI
and CD – if it’s driven on Automation, Lean, Measurement
and Sharing – then as an industry we need to focus on
getting the foundations right before we start certifying the
outcome. I would much rather see training effort put into
Automation tools, understanding of lean / agile practice
and so on.”
Contino’s Benjamin Wootton takes a more conciliatory
tone when he writes “I’m probably one of the few people
in favour of something like the DevOps Institute. Perhaps
this isn’t perfect in its current form, but it’s a step in the
right direction from what I can see. Why? Because I think
DevOps needs to be made more ‘enterprise friendly’ –
packaged up for that audience and made easier for them to
consume and adopt.
“As a concept, I think this kind of organisation could help,”
Wootton suggests. “A ‘grown up’ organisation like the
DevOps Institute, talking in the right language and bringing
the associated training and certification could remove some
of those unfair objections, give DevOps a level of maturity,
perhaps get it ‘in the door’ of more organisations and allow
the other advocates to follow up behind with the real meat
of the message.”
For its part, the Institute recognises that its existence has
sparked a debate in the DevOps community. In a blog
posting at DevOps.com editor-in-chief Alan Shimel admits
that consensus was unlikely any time soon.
“I know there is a segment of the DevOps community
who view this as an apocalyptic event,” he notes. “While I
obviously don’t agree with that view, I respect their opinion.
As I have said all along the DevOps community is big and
getting bigger. There is room for a wide spectrum of opinion.
“If you don’t think that DevOps needs formal education,
training and certification, so be it. I and many folks I have
spoken to believe that as DevOps continues to cross the
chasm into mainstream enterprise IT, a formal training and
certification program will be required. Whether it is from
the DOI or some other entity is the only question.”
For my part, I do see the need for a body advocating
standards for the DevOps movement as it continues to
integrate with mainstream IT. It could also be a valuable
voice in support of Government and training initiatives,
lobbying for investment in this area - which is what I would
like to see the Institute start to do. However, a stronger case
for certification is needed as the current pass rate of 65% is
fairly low. For example, the UK driving test written (theory)
“The DevOps community
is big and getting bigger.
There is room for a wide
spectrum of opinion.”
devops perspectives 3 | DevOps institutional thinking
19. 19
exam requires 43 out 50 questions answered correctly –
an 86% pass rate. A cynic might suggest that 65% is the
lowest threshold to pass whilst retaining credibility and
maximising the chances of those who have paid for the
course passing and becoming “certified”. You can have a
third of the multiple choice questions wrong and still be
certified; not an especially high bar.
Coming back to the driving test analogy, whilst passing the
written and practical components gives people the certification
to drive a car, it doesn’t necessarily guarantee that they are
competent drivers, nor does it ensure that they will remain
good drivers for years to come. Practical application of what is
learnt on the course and on the job are essential.
To avoid going down a similar route and producing below-
par DevOps practitioners, the DOI may need to revisit its
criteria in the near future to ensure that, as DevOps becomes
more a part of the mainstream IT environment, it is handled
by the best. Finally, one aspect that has surprised me has
been the lack of a “buzz” around the Institute, positive or
negative – tweet and follower volumes have not achieved
any critical mass and response overall seems to be muted.
It will be interesting to revisit the Institute’s progress
12 months from now and witness its take-up, feedback
generated and overall evolution.
devops perspectives 3 | DevOps institutional thinking
“It could also be a
valuable voice in
support of government
and training initiatives”
20. 20
Whatis“Micro”about
a Microservice?
One of the hottest new terms in the world of enterprise computing
is the microservice. Starting with the seminal 2014 article by James
Lewis and Martin Fowler of ThoughtWorks, microservices have taken
on a life of their own – and as with any other overhyped term, they
have generated their fair share of confusion as well.
Perhaps the best definition of microservices comes from Janakiram MSV, Principal at
Janakiram & Associates. “Microservices are fine-grained units of execution. They are
designed to do one thing very well,” according to Janakiram. “They contain everything
from the operating system, platform, framework, runtime and dependencies,
packaged as one unit of execution.” As a result, “a microservice architecture promotes
developing and deploying applications composed of independent, autonomous,
modular, self-contained units.”
And yet, the above definition leaves us with yet another question: what does it mean
for a unit of execution to be fine-grained? The obvious answer is small, as in micro, the
prefix that gave microservices their name. But does this notion hold water?
Jason Bloomberg, President, Intellyx
devops perspectives 3 | What is “Micro” about a Microservice?
21. 21
Unfortunately, granularity is a rather general term. Many
different things can be more or less granular, or finer- or
coarser-grained. The notion of granularity in the context
of services arose in the early 2000s in the context of web
services: contracted software interfaces that comply with
a set of XML-based standards.
However, since web services were interfaces rather
than software components, granularity referred to the
granularity of the interface. In other words, a fine-grained
service sent and/or received a small number of values
(say, a single number or string), while a coarse-grained
service sent and/or received structured information that
contained several values (for example, an XML document),
as shown in the diagram above.
Determining the appropriate granularity for a service
interface was tricky, as there were pros and cons for any
level of granularity. Fine-grained interfaces generally
lacked business context, but tended to be more reusable.
Coarse-grained interfaces, on the other hand, often had a
clear business context, but were typically purpose-built for
a particular situation, thus limiting their reusability.
Microservices, in contrast, are more than interfaces. They
are the whole package, as Janakiram points out—a unit of
execution including code, runtime, and more.
In his new book, Building Microservices: Designing Fine-
Grained Systems (O’Reilly, February 2015), Sam Newman
says, “The question I am often asked is how small is
small? Giving a number of lines of code is problematic for
a number of reasons, including language differences and
the specifics of the task at hand”.
He offers a practical measure. “If the codebase is too big
to be managed by a small team, looking to break it down
is very sensible,” Newman posits. However, he adds a
caveat: “the smaller the service, the more you maximise
the benefits and downsides of microservice architecture.”
The challenge of
defining granularity
devops perspectives 3 | What is “Micro” about a Microservice?
In other words, make your microservices too small and
you’ll have to manage excessively large numbers of them,
but make them too big and you’ll lose the benefits that
drove you to create microservices in the first place.
Newman makes one other comment on the subject:
“Another somewhat trite answer I can give is small
enough and no smaller.” Trite, perhaps, but this point
underscores an important principle of microservice
construction: parsimony.
A parsimonious microservice is as small as it should
be – and as Newman says, no smaller. In other words,
during your iterative refactoring efforts (part of any Agile
approach), revisit your microservices and see if there’s
anything extra in them, or if splitting up a microservice
into two or more microservices improves matters. If so,
continue to pare them down and split them up until
such efforts no longer move your project forward.
Microservice with
Fine-Grained Interface
Single Value
Microservice with
Coarse-Grained Interface
JSON
or
XML
22. 22
Revisiting software cohesion
Another important microservices principle – and in fact, a
principle of modular software design since the 1960s – is
cohesion. A microservice is highly cohesive if its elements
belong together – in other words, it does one thing and
does it well, as shown in the diagram above.
Web Services frequently suffered from low cohesion.
Sometimes WSDL files contained dozens or even hundreds
of operations, where a single service did a wide range of
different tasks. Reacting to the problems with that kind of
service is one of the primary motivations for microservices.
Therefore, while parsimony and cohesion make more
sense in the context of microservices, granularity makes
more sense in reference to interfaces – and the concepts
are quite different. After all, a well-built microservice might
(and typically would) have a coarse-grained interface, for
example, if it accepted and/or returned a JSON document.
devops perspectives 3 | What is “Micro” about a Microservice?
However, as the title of Newman’s book suggests, an
entire system built with microservices would itself be fine-
grained – yet another sense of the notion of granularity. A
collection of microservices would be fine-grained if those
microservices tended to be parsimonious and highly
cohesive – regardless of how big they were, and even
though their interfaces should be coarse-grained.
If it were up to me, I wouldn’t refer to microservices or
systems of microservices by their levels of granularity
at all – but unfortunately, it’s not my call. So heed this
warning: granularity is a slippery concept. Focus instead
on the parsimony and cohesion of your microservices, and
the granularity of their interfaces.
Microservice with
Low Cohesion
Microservice with
High Cohesion
“Volorrum qui
consequat quaepudam,
coritat iatio. Pores estiisi
nonsequi ulluptas ipidel
inveror simolendi sum
eos aut pla dolorro
23. 23
Today, DevOpsGuys is at the forefront of technology
promoting Lean IT, DevOps and Continuous Delivery
across the enterprise to ensure alignment with
commercial opportunities, speed-to-market and client
impact. Based on that experience, Smith perceives a
need to rethink some roles:
“Typically there are a bunch of incentives for DevOps which
are based around driving behavior,” he says, “but we’ve
been focusing on the wrong things.”
Smith lists numerous examples of inappropriate
incentivization in DevOps circles. For example, in some
organizations, developers are rewarded for the number
of bugs that they fix, but this can simply lead to the
release of buggy code: the more bugs released, the more
there is to fix.
Then you’ve got the so-called ‘death march’, where DevOps
people are incentivized by the number of hours that they
work. In these cases it makes little financial sense to finish
as quickly as possible when more hours = more money;
something that should take one hour suddenly takes ten.
Or there’s the expert incentive, where someone is given
ownership of a piece of technology and becomes,
effectively, the single point of failure. What can happen
next is those people end up becoming the only ones who
understand the situation completely. The organization ends
up incentivizing ‘knowledge hoarding’ over the sharing and
distribution of information.
How should DevOps be incentivized? It’s a question that organizations increasingly find
themselves asking. James Smith, one half of DevOpsGuys, has over 15 years’ experience
delivering and managing enterprise web applications for global blue chip companies.
Flipping incentivization
James Smith, Co-Founder, DevOpsGuys
“The more bugs released,
the more there is to fix”
devops perspectives 3 | Flipping incentivization
24. 24
Smith argues for the re-examination of roles and their
responsibilities and the incentives typically applied to them.
“If you’re in development, then you are incentivized to
change things. You’re all about developing new features,
changing existing functionality and so on,” he says.
“So you’re incentivized for speed and risk: the faster you
can get stuff out, the better. But if you’re in operations
then you’re largely incentivized to reduce change, to
maintain stability, certainty and predictability.”
That’s the traditional model for organizations, Smith
believes. To break the mold and better align incentives
across the teams, swap the roles around so that you
incentivize the development team on stability, uptime and
managing certainty, and incentivize operations on value
creation and how well they innovate and manage change.
“You want to make sure that both sides are carefully
aligned and understand what the other does,” he
suggests. “You can put each group in the other’s shoes.
You have them sitting on the other side of fence.
“Take the development team and put them on call
and make them responsible for the software. They
Role reversal
take ownership of front-line issues, dealing with the
support tickets, which should give them an idea of
the supportability of their software and a greater
understanding of what they’re producing.
“In the same vein, the ops guys are pushed into the
development process, so they get an understanding of the
best practices of managing change, having to use source
control, peer-reviewing code, looking at automation
techniques and so on. This should give them an idea of
how they can enable change quickly in their organization.”
Having this shared perception of one another’s roles
enables a realignment of goals around both groups, but
in this process it’s important to bear in mind that the one
goal absolutely not up for discussion or compromise is
quality. “Both teams need to be centrally aligned around
customer value and quality,” cautions Smith. “When the
goals are aligned, the rate of delivery and the quality of
the product increase.”
All of this definitely needs senior stakeholder support
within the organization and you’re almost certainly going
to run into some ground-level resistance to change. Smith
notes: “People are creatures of habit. They’re brought in
devops perspectives 3 | Flipping incentivization
“When the goals are
aligned, the rate of
delivery and the quality
of the product increase.”
25. 25
to do a role, then you ask them to do something else and
they say ‘That’s not in my employment contract’.
“You’ll find that in some companies there is resistance
of that nature, but it’s really fear of the unknown, so you
need to show the benefits and help them to see that the
changes are going to help them to do their jobs better.”
Ration these role-reversal periods as well. “Initially this should
be done for short periods of time,” says Smith. “But the
frequency is important. It shouldn’t be a one-off exercise.”
It’s also essential to be able to measure the impact
of the role changes. “You can start to encourage new
behaviors, but unless you measure the impact you’re
just dealing with opinions, not facts,” warns Smith. “Get
measurements around the rate of change or get quality
metrics or whatever you want to measure. But do put
some kind of measurements in place.”
Overall you need to be able to show that adjusting the
traditional incentivization practices helps DevOps people
to do their jobs better and contributes to the benefit of
the organization.
“At the end of the day, you are improving your application
quality. You are improving the products that you are
producing,” concludes Smith. “There’s a strong correlation
between creating high-performing IT teams who are able
to deploy products more frequently and keep uptime at a
higher level and contribution to the bottom line.”
devops perspectives 3 | Flipping incentivization
26. 26
From a business perspective, she’s had good and productive
engagements with IT counterparts, but equally she’s had
experiences that were less so, at times bordering on the
hostile. After discovering Agile, she became a passionate
advocate of its ability to align business and IT around the
delivery of value and of the potential of introducing DevOps
to enterprise organizations.
The first thing the business decision-maker needs to do, she
advises, is get informed and assume responsibility for your
own education. Don’t assume you’re going to get all the
information that you need unless you ask for it.
She says. “I spent a very long time as a business person and
as a business person you just assume that the IT folk know
what they’re doing. Furthermore, we tend to be conditioned
not to question that assumption.
“When I started learning about DevOps and Agile, that
was when I started asking questions. You really do need
to ask questions. You need to read the books and attend
the conferences and ask the questions because that is
how you move from the assumption that all is well with
the technical practices within your organization to a more
realistic understanding of the true state of play.”
In a career spanning more than 20 years in business, Em Campbell-Pretty, Partner at
Australia’s leading enterprise Agile consultancy Context Matters, has witnessed first-hand
some of the tensions involved in that long-standing dilemma: how to bridge the gap
between business and IT.
Educating DevOps
Em Campbell-Pretty, Partner, Context Matters
devops perspectives 3 | Educating DevOps
27. 27
Campbell-Pretty is a firm believer that it’s incumbent
on business people to get up to speed, if only to put
themselves on a level playing field with IT. “I had an
Agile coach who believed that people should educate
themselves. That pushed me to focus my reading time
on Agile classics like Jim Highsmith’s Agile Project
Management,” she explains. “It’s just so, so powerful just
to take the time and investigate.
“It’s not easy to get people to read, but those who do
are far more successful. I recently heard Verne Harnish,
author of Scaling Up, speak at a conference and he made
the point that ‘those who can read, and don’t, are only
marginally better off than those who can’t’ and he’s right,”
she adds.
“If nothing else, reading up enables you to ask questions
of the right people. I remember asking IT what the
Educate yourself
automated test strategy we had was and getting a blank
expression back. The question hadn’t been expected. The
IT people didn’t like that much.”
That blank look reminds us that there’s obviously a
danger that challenging IT on their home turf might not
be welcomed, but it’s worth it, reckons Campbell-Pretty.
“It doesn’t always go down particularly well with IT when
the business starts asking questions,” she admits. “But
by the same token the IT folk don’t always take time to
explain everything to the business.”
Campbell-Pretty has personal examples that illustrate
her point. “The first time I picked up a book on Agile, it
was because I’d had a conversation with an IT project
manager who bombarded me with jargon, she recalls.
“The project manager came to me and said that she
intended to deliver the project using Scrum, which would
involve locking people in a room for a month, at the end
of which they would emerge with working software. My
head was spinning. What nonsense were these people
talking about? So I logged on to Amazon and ordered
some books on Agile.
“A couple of years later, the role of business sponsor for
that particular project was transitioned to me. The project
had not ended up using Agile or Scrum and delivery was
not going well. We were spending a lot of money and
not getting business outcomes. There had to be a better
way. This brought me full circle back to the reading I had
done about Agile and eventually the entire program being
transitioned to use Agile and later including DevOps.”
devops perspectives 3 | Educating DevOps
28. 28
But if the business side is getting armed with the right
questions to ask, the next question is how far CIOs are
themselves up to speed with the answers, particularly
around DevOps. Campbell-Pretty is uncertain on this point.
“It’s hard to generalize if CIOs get DevOps. I don’t think it
has gone mainstream yet,” she suggests. “There are very
big companies out there which are playing with DevOps,
although it seems to be mostly in the digital space. Taking
on DevOps in large companies with big, heavy, legacy
applications is a very different challenge. I’m not sure how
many enterprises are facing into the DevOps challenge at
the moment.”
In fact, Campbell-Pretty makes the case that CIO
involvement may not always be entirely helpful to the
cause of successful DevOps introduction to an enterprise.
“What tends to happen when the CIO gets involved is that
the organization sets up a DevOps Centre of Excellence,”
she argues. “Frankly I’m not sure that I buy into that as a
CIO – help or hindrance?
mechanism for rolling out DevOps because it becomes
a top-down mandate. What you need to do is to change
the culture and that’s just not the same thing as someone
swooping in from head office and saying ‘Do what I say’. It’s
in no way as powerful as a ground-up approach to DevOps.
“You do need funding and expertise and those will likely
come from a central source, but when these things
become CIO-driven, you can find yourself with a problem.
If you’re going to try to solve the DevOps problem in large
organizations through a central mandate, then you’re
going to miss out on huge opportunities for people to
get better and better by improving what they do day in
and day out. You need to create tribes and harness their
energy to inspect and adapt and innovate.”
At the end of the day, it’s essential to get those business/
IT conversations happening for mutual benefit—and that
involves both parties upping their mutual respect. “We
have a problem that there is a lack of business people who
respect technology people and vice versa,” she concludes.
“The right conversations just don’t happen enough.
“Business people are pretty bright you know. They run
million-dollar businesses. So IT really can go to business
people and talk to them in science-based and fact-based
terms. Help them understand why it makes sense from a
business perspective to invest in DevOps. Then they will
get on board and potentially even become your greatest
ally in your journey to DevOps. ”
devops perspectives 3 | Educating DevOps
“Do what I say’ is in no
way as powerful as a
ground-up approach to
DevOps.”
29. 29
However, before automating it is essential to consider its
opportunity cost, the cost of everything else you could be
doing instead, and whether automation is even appropriate.
Specifically, is automation being applied to the right things,
at the right time in the project?
When delivering software, automation
seems like a really cool trick. You can easily
show a manual release or build, do some
automation ‘magic’ to show that the build is
now automated, put metrics on it, show it is
more efficient and deliver huge time savings.
Doing that can be very popular.
Wielding the
double-edged
sword of
automation
Dan North, Dan North Associates Ltd
devops perspectives 3 | Wielding the double-edged sword of automation
30. 30
If you have a deterministic transformation, such as
compiling code to binary, where all of the stages are
tightly defined, this is clearly a good candidate for
automation. This is why hand-compiling code is a
minority sport now.
However, automation becomes a double-edged sword
when it automates a process or activity, as these tend to
be a function of context. If that context drifts, then the
automation solution may no longer be appropriate, and
may even be detrimental.
For example, with cars it currently makes sense to
automate transmission but not steering, as steering is
a function of the behavior of other cars on the roads.
Although, as Google and others have shown, you can
manage to automate steering, even around ducks, it is
just not currently economic to do so.
The choice over braking automation is, however, more
clear-cut. You can teach drivers that in an emergency,
cadence braking will let them steer and slow at the same
time, but the reality is that in an emergency most drivers
will fail to do so. An ABS braking system will ‘remember’
to cadence brake, and do so more effectively, every time.
This makes applying automation here deterministically a
better solution than the manual alternative.
The early stages of a project are when you know
the least about everything—the organization, the
technology used, the operations environment, the
Choosing what
to automate
constraints and your team—yet this is often when
people reach for that automation ‘magic’, seeking
speed and repeatability. As soon as the team has
what looks like working code, they automate the build.
Unfortunately, this crystallizes the current knowledge of
the project, including all your erroneous assumptions.
Any errors in understanding of how things work are
baked in, and the chances are they will stay there and
never be reviewed. They remain fixed by the ‘we have
features to ship’ mentality, no matter how slow or
complex the build is, it is THE build and there is always
something more pressing to deal with.
One of the core principles of Agile is adapting to
change over following a plan. So automation, which is
mechanistically following a plan, needs to be challenged
and adapted as part of Agile working. Go back to manually
building if necessary, question the sequence of events,
evaluate options for parallelization, try running slow
processes first, test whether the impossible can happen,
as what was once best fit is often no longer optimal.
devops perspectives 3 | Wielding the double-edged sword of automation
“Fugitae as ipis
alit eictate verfers
perciatem es quodis”
Visiting a team working for a financial services client,
I found them investing enormous effort constructing
their build pipeline. It was a thing of beauty, with
automated testing stages, reporting and all kinds of fancy
instrumentation. I visited them around six weeks into the
project and I asked what demonstrable client features
they had delivered. They looked embarrassed—nothing
had gone through that pipe.
The conversation with the client was equally
uncomfortable. They had bought into the benefits of
automation but to their stakeholders this was looking
anything but agile.
Getting the process of automation right requires
considering what you are not doing while you are
automating, the opportunity cost and where the value is
for the client.
The opportunity cost
of automation
31. 31
So how do you discover where automation will add
the most value? I advocate recognizing that there are
different types of work that are valuable to a project.
Agile methods like Scrum tend to only explicitly
recognize one type of work, which is delivery of features.
Fundamentally, all Scrum measurements are features
metrics: velocity, story points, burn-up, burn-down. As
well as feature delivery, I believe teams should also value
two other types of work as first-class citizens, namely
discovery or exploratory work and Kaizen.
Discovery work, understanding more about the problem
you are trying to solve, is first-class work. It happens
anyway, it is what people do, but it is not explicitly
recognized and so is undervalued. Active discovery can
be the key to a much shorter path through the problem.
Maybe you can get the same business outcome with
fewer, different features.
Kaizen in this context is not just confined to the narrow
definition of continuous incremental improvement, it
encompasses improving the system in which you are
operating. Knowledge transfer is a great example of
applying Kaizen to the delivery system. If one team
member has a skill and teaches that skill to another,
capacity for that work is doubled. If tacit knowledge is
documented, it becomes available to all the team, which
Valuable work
increases their capacity to solve problems, so we have
made a better system for delivery.
Done well, automation is a form of Kaizen. It improves
the development process, speeds testing and reduces
likelihood of defects by eliminating manual work. But
since Kaizen is rarely recognized as first-class work it gets
done ‘in the cracks’ so it is not governed, not subject to
the due diligence and oversight of delivery work. There are
small, tactical instances of explicit Kaizen activities, such
as during a ‘sprint zero’, where the build gets automated,
but as discussed earlier this is often the wrong time.
So Kaizen gets hidden. Developers instinctively know it
is useful, and indeed necessary, to sharpen the axe, but
with no ‘official’ sanction to hang the work on they end up
flying under the radar to do it anyway.
devops perspectives 3 | Wielding the double-edged sword of automation
Investing in discovery and Kaizen work is an example of
having made the choice between hacking and strategic
development. It needs to be elevated to first-class work
and made visible and demonstrable.
Demonstrating delivery is simple: you can demo a new
feature, showing something that was not there before.
The only way to ‘measure’ discovery or research activities
is by time-bounding. Allocating time to dig into a
particular facet of the project, running experiments,
reviewing data, then asking the question: is there more to
be gained through additional research, will we progress in
this area, or are returns diminishing?
We can demonstrate knowledge transfer using a show-
and-tell by the newly trained team member, or by having
them document their learning.
Kaizen for the delivery process can be measured in
terms of how much time will be saved next month by
automation, e.g. time not spent debugging or diagnosing
flaky manual deployments, and can be assigned a value
just like features.
“Boring is a necessary, but
not sufficient, condition
for automation.”
32. 32
Experience has moved me from ‘If in doubt, automate’ to
‘Don’t automate until something is boring (and even then
maybe not)’.
Boring is a necessary, but not sufficient, condition for
automation. If a process is boring it means you have done
it often enough that you know how it works, and that it
has become repeatable enough that surprises are unlikely.
When you have both of these conditions, you have a
candidate for automation.
With that candidate lined up, you then need to think like
your customer, and assess the relative value of investing
in each type of first-class work.
Apply this discipline and the double-edged sword of
automation can cut you free you from the boring stuff and
let you carve through what matters.
Seek boredom and
customer value
devops perspectives 3 | Wielding the double-edged sword of automation
33. 33
Paul Speers is the CEO of Speerhead. He started the
company over five years ago to direct Speerhead on
the path to long-term growth, further its commercial
success in the DevOps market and drive the launch of
its revolutionary DevOps recruitment franchise and play
a leading role in the creation of DevOps Training and
Certification IP.
Paul brings to Speerhead over 20 years’ experience in
sales and marketing within the IT industry, having held
senior positions at Opsware – the first IT Automation
vendor from Marc Anderson. He is also the co-founder of
Fox IT, the Global ITIL vendor.
Paul Speers
CEO, Speerhead Group
Patrick Hyland is the founder of DevOps Associates, a
London-based consultancy concerned with application
engineering management. The consultancy applies
a blend of agile methods, connected ITIL lifecycle
processes and DevOps collaboration/engineering
practices to help companies design, build, deliver and
operate outstanding application services.
Patrick is an ITIL expert with 18 years of development
and operations experience. He is particularly interested
in management via Eli Goldratt’s theory of consultants,
applying a lean manufacturing mindset within an IT
Service Management context.
Patrick Hyland
Founder, DevOps Associates
Tony Chapman has founded and built specialist DevOps
agency LinuxRecruit who are working with organisations
across the UK, designing their DevOps recruitment
strategy and fully staffing their DevOps Engineering
teams. Tony has been working in the Open Source
community for 10 years and was recently shortlisted for
Recruiter of the Year 2012 at the prestigious national
Recruiter for Excellence Awards, LinuxRecruit were
shortlisted for newcomer agency of the year at the
2013 awards. He is a contributor to the Open Source
community, has a regular column in Linux Format
Magazine and is co-organiser of the world’s biggest
monthly DevOps meetup, the DevOps Exchange in
London.
Tony Chapman
Managing Director,
Linux Recruit
devops perspectives 3 | contributors
34. 34
Matthew Skelton has been building, deploying, and
operating commercial software systems since 1998.
Co-founder and Principal Consultant at Skelton Thatcher
Consulting, he specialises in helping organisations
to adopt and sustain good practices for building and
operating software systems: Continuous Delivery,
DevOps, aspects of ITIL, and software operability.
Matthew founded and leads the 1000-member London
Continuous Delivery meet-up group, and instigated
the first conference in Europe dedicated to Continuous
Delivery, PIPELINE Conference. He also co-facilitates the
popular Experience DevOps workshop series and is
co-editor of Build Quality In, a book of Continuous
Delivery and DevOps experience reports.
Skelton Thatcher Consulting
Grant has driven real collaboration between Operations
and Development teams in AOL, Electronic Arts and
British Gas by implementing Infrastructure as code and
driving application integration from continuous build
systems. Grant has delivered game platforms running
in the cloud enjoyed by millions of players per day and
websites serving a billion page views per month. Most
recently he has delivered a high performance, scalable
Internet-of-things platform for British Gas. Grant is
the author of Next Gen DevOps: Creating the DevOps
Organisation and is frequently sought out for his cloud
and DevOps expertise. Grant can be reached at
grant@nextgendevops.com
Matthew Skelton
Co-founder and Principal Consultant,
Skelton Thatcher Consulting Ltd
Grant Smith
Author, Next Gen DevOps
Nicole is the Director of Organisational Performance &
Analytics at Chef and an Assistant Professor of MIS and
Accounting at the Huntsman School of Business at Utah
State University. She received her PhD in Management
Information Systems and her Masters in Accounting
from the University of Arizona. She is an expert in
IT use, DevOps impacts, and communication and
knowledge management practices, particularly among
technical professionals. Her background spans analytics,
enterprise storage (specialising in RAID performance),
cost allocation, user experience, and systems design
and development. She is a featured speaker at industry
and academic events and is involved in women in
technology initiatives.
Nicole Forsgren
Director of Organizational
Performance and Analytics, Chef
devops perspectives 3 | contributors
35. 35
James, co-founder of DevOpsGuys, has over 15 years’
experience delivering and managing enterprise
web applications for global blue chip companies.
Today, DevOpsGuys are at the forefront of technology
promoting Lean IT, DevOps and Continuous Delivery
across the enterprise to ensure alignment with
commercial opportunities, speed-to-market and client
impact.
Justin Vaughan-Brown is Global Digital Transformation
Lead, Product Marketing at CA Technologies. He is the
author of ‘The Digital Transformation Journey: Key
Technology Considerations’ paper, hosts the quarterly
DevOps Influencer Dinners and is responsible for the
DevOps Simulation Experience, an interactive online
workshop that explains core DevOps principles.
James Smith
Co-Founder, DevOpsGuys
Justin Vaughan-Brown
CA Technologies
Jason Bloomberg is the leading industry analyst and
expert on achieving agile digital transformation by
architecting business agility in the enterprise. He writes
for Forbes, Wired, and his biweekly newsletter, the
Cortex. As president of Intellyx, he advises business
executives on their digital transformation initiatives,
trains architecture teams on Agile Architecture, and
helps technology vendors and service providers
communicate their agility stories. His latest book is The
Agile Architecture Revolution (Wiley, 2013).
Jason Bloomberg
President, Intellyx
devops perspectives 3 | contributors
36. 36
Em is a Partner at Context Matters, Australia’s leading
Enterprise Agile consultancy. After close to 20 years in
business management roles within multinational blue
chip corporations, Em discovered Agile and became
passionate about the chance it provides to align
business and IT around the delivery of value. In 2012,
she launched Australia’s first Scaled Agile Framework
(SAFe) Agile Release Train. Em is an active member
of the global agile community and was invited to co-
chair the Enterprise Agile track for the Agile Alliance
conferences in 2014 and 2015. Em also blogs about her
“Adventures in Scaling Agile” at PrettyAgile.com. Em can
be contacted at em@contextmatters.com.au
Em Campbell-Pretty
Partner, Context Matters
Dan North uses his deep technical and organisational
knowledge to help CIOs, business and software teams
to deliver quickly and successfully. He puts people
first and finds simple, pragmatic solutions to business
and technical problems, often using lean and agile
techniques. With over twenty years of experience in IT,
Dan is a frequent speaker at technology conferences
worldwide. The originator of Behaviour-Driven
Development (BDD) and Deliberate Discovery, Dan has
published feature articles in numerous software and
business publications, and contributed to The RSpec
Book: Behaviour Driven Development with RSpec,
Cucumber, and Friends and 97 Things Every Programmer
Should Know: Collective Wisdom from the Experts. He
occasionally blogs at http://dannorth.net/blog.
Dan North
Dan North Associates Ltd
devops perspectives 3 | contributors