Understanding DevOps – A Perspective

Over the years, DevOps has brought about a welcome change in the way product development and operations teams work and collaborate while paving way for faster time-to-market and enhanced customer satisfaction levels 

Did you know that DevOps is as much about culture as it is about digital transformation /automation?

You read it right.

The underlying current for a successful DevOps practice is the culture of experimentation, shared learning and collaboration at its best - between product development and operations teams. Interestingly enough, this has paved the way for a new and better way of the software development process.

So then, how do we define DevOps? 

Since DevOps has a lot to do with culture and not just technology, there can be many definitions and here are a few:

DevOps is the practice that brings together the participation of development engineers and operations admins with an objective to enhance and expedite the application management lifecycle, from design through to development, testing, deployment, and monitoring. (Source: The AgileAdmin)

DevOps stands for a change in the IT Culture where the focus is on achieving rapid IT product/service delivery through the active collaboration of the development and operations teams using DevOps practices such as Continuous Integration, Continuous Delivery, and Continuous Monitoring. (Source: NewRelic)

DevOps is defined as a set of practices that allows for automation of processes between the software development and operations teams such that software applications are built, tested and released into the market more reliably and quickly. (Source: Atlassian)

The DevOps model is a continuous loop of active collaboration between the development and operations teams such that efficiency is achieved even as time and efforts are minimized.

Some core DevOps principles –

• Flow - that accelerates the work from development to operations to customers
• Feedback – that forms the backbone of the practice and enables the creation of better product
• Continuous Learning - that helps enhance the skill sets of DevOps engineers and redefines the way teams operate
• Automate - the developmental tasks, and the entire IT infrastructure landscape over time

 Why DevOps matters? 

The first thing about DevOps practice is the shift in culture. It is a shift from the regular siloed functioning to a more participative and inclusive style. This is a welcome change for organizations.

So why does DevOps matter, more than ever? 

“   Organizations that are best able to leverage agile and DevOps consistently, have seen a 60% higher rate of both revenue and profit than their mainstream peers who have not” 

• CA Technologies Survey (Nearly 1,300 IT and Business Leaders)

The world of business is going through a rapid change all thanks to the massive proliferation of the internet and the influence of software tools.

Initially, software functioned as an accessory to business operations but that is changing now; it’s fast becoming an integral part of a business. Under these circumstances, there is an imminent need to not only develop quality software but update it periodically and deploy it in the quickest possible manner. This is where DevOps comes into play - it just changes the way software applications are built, tested, deployed and monitored.

The software has become ubiquitous and essential. From bridging customer communication gaps to increasing operational efficiencies and delivering greater customer experiences, it plays a major role in business operations. Software applications help businesses put their best foot forward and DevOps helps build quality applications and at a faster rate.

DevOps Market Size

As per a MarketsandMarkets report, the global market for DevOps was about $ 2.9 billion for 2017 and is projected to grow at a CAGR (Compounded Annual Growth Rate) of  24.7% over a five year period to touch $ 10.3 billion by 2023. Globally, while North America will account for the largest share in the DevOps market, Asia Pacific will account for the highest growth rate during the said period.

The above stats only reaffirm the fact DevOps is fast being embraced as software development methodology by businesses across sectors.

The DevOps Life-Cycle

The DevOps lifecycle is all about getting two teams to come together to become ‘’one’’. You might be wondering what we talked about collaboration and not unification. Well, the core philosophy of DevOps is that development and operations teams collaborate so well to function as one big extended team. The life-cycle is all about shared experimentation, learning and quality product development.

‘’DevOps is not a Goal, but a never-ending process of continual improvement’’  

- Jez Humble, Author – The DevOps Handbook: Lean Enterprise

Let’s look at the various stages of this life-cycle -

Plan: It all starts with planning - planning about the kind of application or software that needs to be developed. This involves making a schematic about the requirements and development process.  

Code: Coding is the beginning of the execution process towards building the software application based on the client’s requirements and proposed plan of action.

Build: Various codes delivered in the previous stage are then used to build an application. Here, application development is broken into multiple ‘’sprints’’ that have shorter development cycles.

Test: This is an important phase for it helps developers comprehensively test the application for bugs and fixes. It is the phase that validates coding and building processes and allows for the release of the application.

Release: The tested application is then released into the market and is, Live. Sometimes, companies release the newly developed application or functionality to a test audience, other times, it is directly released to the entire market.

Deploy: This is the phase where the code is deployed either into an on-premise or cloud environment. Deployment is performed in a manner that does not interfere with the functioning of the application.

Operate: Here, operations are performed on the code as and when needed.

Monitor: The released application is monitored for its performance. The performance is noted and key changes (to be made) are listed as per client requirements. These key changes then translate into further requirements wherein we again move to the planning stage initiating what is a continuous loop between development and operations teams.

In simple terms, DevOps Life-cycle combines the functions of development and operations teams to roll out a better product based on client requirements and feedback in a quicker time frame.

DevOps Practices

What really differentiates DevOps from other methods of software development is the continuity of learning and application of this learning to build a better product every time. The elements or best practices of DevOps are what make it stand as a software development methodology that is finding great acceptance across industries.

Let’s look at what these best practices are and how they help -

• Continuous Integration (CI):

It isn’t enough if a new code is built, it needs to be integrated into existing code or application. CI aims at integrating codes built by developers into the central code repository more frequently such that detection of bugs, if any, can be carried out more quickly.

New code may also represent a new functionality to the existing application in which case it has been integrated into the existing code as smoothly and as quickly as possible - Continuous integration is the way to do this.  Another thing to keep in mind is that the changed code should not, as much as possible, allow for any errors in the run-time environment.

The ultimate objective of CI is to make code integration repeatable and easy so as to reduce overall costs and quickly discover challenges at integration.

• Continuous Development:

This is the stage that differentiates the traditional waterfall method with DevOps method – the software or application is developed, continuously. Here, the software development process comprises of multiple ‘’sprints’’ or phases of development cycles that are short and are delivered in a faster time frame. This is the stage where the developers’ technical skills are tested as they focus on coding and building.

• Continuous Delivery (CD): 

CD is an ongoing process where the code is built, tested, configured and deployed to a production environment. This follows upon the CI stage; all code changes are deployed to production environment after the build stage. A characteristic feature of CD is that the code is always in a ready-to-deploy state.

What is central to CD is the continuous delivery pipeline – a set of processes that make use of tools to compile, test and deploy code for additional/new features.

• Microservices: 

Microservices is a valuable approach to application development wherein a single large application is built as a suite of smaller services. Each service is independent of the other while it is connected to the others through interfaces. Another advantage of Microservices is that problems in one service do not impact other services thereby reducing the number of failures in development. The commonality between Microservices and DevOps is that both advocate decentralization of work process allowing for small teams to decide on their delivery.

CI and CD can be effectively used to drive Microservices that increase development velocity.

• Infrastructure as a Code (IaC): 

In simple words, IaC is the practice of managing the operations environment in the same way as it is done for a code or application. Configuration changes are common as part of infrastructure management. Manual configuration of infrastructure consumes a lot of resources and time not to mention the high possibility of human errors. DevOps eliminates the issue by helping to switch from manual configuration to automatic or programmatic with the help of continuous integration, continuous monitoring, and version control. IaC also helps in the code review process; there is much clarity and teams have a clear idea on the changes so everybody is on the same page.

• Monitoring and Logging: 

Monitoring and logging form an integral part of quality product delivery and eventual customer satisfaction. They help organizations understand how the performance of an application impacts customer experience. Data and logs from application usage are captured and analysed to understand how changes impact product quality and customer experience. Active monitoring is a must even as changes to IT infrastructure and application become more frequent.

• Communication and Collaboration:

 No matter how good technology or a methodology is, nothing can be achieved without human interference. The cultural takeaway from DevOps is that it enables better communication and collaboration between two teams without which the DevOps practice collapses.

Open communication and collaboration are what will lead to enhanced learning experiences and improve combined productivity. In fact, communication and collaboration are even extended to other functional teams such as Sales, Marketing, Finance, etc. bringing a paradigm shift in the way an organization functions towards achieving its objectives. Now, isn’t that a great advantage of DevOps practice?

 DevOps Tools and their relevance

For DevOps practices to succeed, it needs tools and excellent tools at that. As we move into 2019, here are a few major DevOps tools -

• Git – The DevOps practice usually starts with a tool like Git. It is a version-controlled system used for tracking changes to the code while developing the software application. It is also used to track changes in the set of files and also helps in coordination among programmers.
• Jenkins – A popular Continuous Integration tool, Jenkins is an open-source automation tool written in JavaScript. It helps to automate non-human tasks of development and is easy to install and maintain. The other popular CI tools are – Apache Gump, Bamboo by Atlassian and CircleCI.
• Selenium– It is a portable and open-source framework/tool that is used for testing web applications. Selenium is a suite of software, each of which caters to a different need. There are four components to the tool – Selenium Integrated Development Environment, Selenium Remote Control, WebDriver, Selenium Grid.
• Docker – Arguably the most popular container platform, Docker is open-source and helps teams build, ship and run distributed applications. It helps users assemble apps from different components and work in collaboration. It also helps to schedule and orchestrate containers on machine clusters.
• Kubernetes – A widely popular tool, Kubernetes is a container orchestration system and a game-changing tool. It helps in running containers in a cluster. It also helps manage and monitor different applications, effectively. Kubernetes is an open-source system originally built by Google and is now being managed by Cloud Native Computing Foundation or CNCF.
• Chef and Puppet – As two top configuration management tools, Chef and Puppet have a lot in common and are used for deploying, configuring and managing servers. They are both open-source and are based on the master-slave architecture. The major difference being - Chef is largely considered a ‘’Developer’s tool’’ allowing for more creativity and is easier to deploy and scale. Puppet, on the other hand, is more of a ‘’safe’’ tool that allows less liberty. Other popular configuration management tools are Terraform, Ansible, and Saltstack.  
• Nagios – An open-source and free system monitoring tool that was designed to run on Linux Operating System. It can help monitor devices that run Linux, Unix, and Windows. Nagios helps identify and resolve IT infrastructure issues and thereby helps avoid critical business process problems.
• ELK Stack – A combination of three powerful tools – Elastisearch, Logstash and Kibana, ELK Stack provides a value add as it helps collect and analyse logs to provide some great insights. It is an open-source tool with free plug-ins.

You can also learn more about DevOps tools here.

How DevOps Works

In purely technical terms - DevOps is a move away from the ‘’Water-fall’’ model of software development and is a more enhanced process than the ‘’agile’’ method. We shall talk about the major differences between DevOps and Agile a little later. For now, let’s understand how DevOps works -

The communication gap between development and operations team is bridged through DevOps workflow. Due to this, siloes are broken giving rise to greater collaboration, work efficiencies, and better product delivery. This ultimately leads to faster time-to-market of products and higher levels of customer satisfaction giving organizations the much-required competitive edge.

At the core of DevOps Lifecycle is the ‘’Continuous’’ philosophy – a simple yet profound philosophy that keeps the DevOps Life-cycle running and improving at each stage.

Benefits of DevOps and its Business Value  

There are several benefits to adopting DevOps. We have observed some of these in the earlier sections, let us now look at the benefits that provide business value to organizations -

• Stable Environment:

DevOps involves breaking down silos leading to better communication and collaboration between the two teams. This leads to a more stable operating environment, a much-required move away from the ‘’chaotic’’ environment that we have been used to.

• Increased Efficiency: 

Stable operating environment automatically brings about increased efficiencies of individuals and teams since it less about competition and proving the other wrong and more about achieving set objectives, together.

• Higher Productivity: 

Efficient teams end up building better quality products leading to lower failure rates, faster fixes and quality time focus on core functions. DevOps teams end up being more productive and less obstructive. This leads to a reduction in CAPEX and OPEX.

"There is an average of 19% increase in revenue directly attributed to adoption of DevOps methodologies. Automating software benefits organizations in multiple value areas, financially’’ - CA Technologies Survey

Faster time-to-market: Arguably, the bigger advantage of DevOps is its ability to deploy software tools or applications in a faster time-to-market manner. This gives an edge over the competition while ensuring that the customer/market feedback is quickly translated into the enhancement of product(s) resulting in customer satisfaction.  

Better ROI: Better products lead to greater customer satisfaction which can be the foundation layer for customer loyalty. Customer loyalty coupled with lower developmental and operational costs is a blessing for any organization as it ramps up the ROI. Over time, this can lead to increased revenues and profitability.

Overall, DevOps just about increases the velocity at which software can be developed, tested and deployed to meet client requirements, successfully.  This is a great value add, especially so, in the current times where customer demand is high and the need to have a competitive edge has gained paramount importance.

DevOps Trends for 2019

As per a Statista report, DevOps adoption has shown an increase of 17% in 2018 vs. only 10% in 2017. The year 2019 will an interesting one for DevOps as more organizations prepare themselves to embrace the practice in a bid to enhance business operations.

In that vein, let’s observe some of the top DevOps trends for 2019 –

• Increased Focus on CD: 

While Continuous Integration is an essential component of DevOps practice, the focus will shift more towards Continuous Delivery where organizations will look to automate the entire software development process to the extent possible. At a macro-level, leadership is looking towards CD as the key component to deliver better deliverables and thereby positively impact business performance and customer engagement in a more comprehensive manner.

• Growth in Microservices: 

Microservices will see compounding growth, there is no doubt about that. As Microservices do not create any dependencies in software development they augment the DevOps philosophy; both go hand in hand. Moving to Microservices architecture helps improve delivery and increases runtime, something, every organization looks forward to in a highly competitive market scenario.

• Data Science and AI integrate: 

Applications generate a lot of data and DevOps is becoming more data-driven these days. With so much data available, it is only prudent for businesses to harness the same to enable better outcomes. Going forward, Data Scientists and AI experts will be working closely with DevOps professionals to observe, study and analyse different trends as users interact with applications. Insights can be gathered to help enhance the DevOps process and also better understand customer behaviour patterns.

• Kubernetes to evolve and lead:

Kubernetes has slowly become the fastest growing technology, and as per reports, it is the highest adopted container technology. Container Orchestration is used by software teams to manage, control and automate various tasks. Each service broken down from the larger application (via Microservices architecture) can be stored in a container. Kubernetes helps to adhere to the CI and CD processes because it is easy to scale, use and manage.

• DevSecOps gains focus: 

Security, in software terms, is usually treated as an afterthought. In fact, this attitude has led to further vulnerabilities in technology adoption and usage. However, 2019 will see a different outlook even as security becomes an integral part of software development – it will be integrated into the development process and not added on top of it. DevSecOps is the practice of putting security first and injecting the same into the development life-cycle of an application. This will gain much momentum in 2019.

What DevSecOps also does is that it makes everyone involved as a stakeholder in the security maintenance process and thus, helps reduce the confusion on how to go about implementing the DevOps Security process.

Is DevOps related to Cloud Computing?

Yes.

While DevOps is about the process and process enhancement, Cloud Computing is all about technology and services. The two are mutually exclusive and yet the common binding factor is that both have become very essential for organizations moving towards digital transformation.

How related are these two? 

First and foremost, the centralized nature of the cloud provides DevOps with the right platform for building, testing, deployment, and production of code. Cloud platforms support DevOps automation; they provide the right environment and tools to ensure continuous integration and continuous development. This allows for lower spends on software development in comparison with on-premise development.

Cloud Computing allows developers to enhance their efficiency and productivity. It also allows them to have more control over their work process resulting in smaller wait times which is a great advantage. Since Cloud provides application-specific infrastructure, developers can now own more components. The various tools available on Cloud help quicken the development process, ensure repeatability and reduce human error to a large extent.

There is a reduction in infrastructure costs as DevOps in combination with Cloud, allows for scalability in the software development process. Cloud helps developers in creating a self-service method as far as provisioning infrastructure is concerned. Developers need no longer depend on IT operations as they can experiment (build, test) and come out successful with quality products.

To sum it up, DevOps and Cloud have a synergetic relationship whose combination can provide multiple benefits to organizations and help speed up their digital transformation in sync with their business objectives.

What is the difference between DevOps and Agile? 

As interesting as it can get, there have been enough and more debates and discussions on DevOps and Agile with respect to software development and Digital Transformation. Let’s understand some basic differences -

While both DevOps and Agile are involved in the creation of software applications, they have clearly demarcated differences.

• Teams involved:

Agile is largely restricted to the development team and involves the continuous iteration of development and testing of code/application in the software development life-cycle. DevOps, as we had earlier discussed, involves both the development and operations teams. It is a culture that promotes greater collaboration and work efficiencies between the two teams.

• Task Focus: 

Agile focuses on managing complex projects by breaking them into smaller pieces with the help of frameworks such as Scrum, Sprint and Safe. DevOps goes beyond to manage the entire engineering process that involves development and system admin teams.

• Task Timeframe: 

Agile tasks are carried out in short cycles called ‘’Sprints’’ and the duration for each sprint can be anywhere from 2-4 weeks. DevOps also has a short development life-cycle, the difference being code is delivered to production on a daily basis.

• Feedback Source: 

While Agile teams obtain feedback from the customers, DevOps teams obtain the same from Operations.

In other words, DevOps is nothing but a continuum of Agile methodology that goes beyond software development team to operations and focuses on faster time-to-market of product and end-to-end business solutions. It is also interesting to note that DevOps facilitates agile development teams to implement the continuous integration and continuous delivery processes resulting in a faster product release.

What are the roles and responsibilities of a DevOps Engineer? 

A good DevOps engineer is in demand and is almost always busy. Yes, good DevOps engineers get paid well. Let’s now, observe some roles and responsibilities of a DevOps Engineer -

• First and foremost, a DevOps engineer must understand the requirements of the client. Not understanding can completely set the task journey of the engineer in the wrong direction.
• Perform problem-solving across application domains and system troubleshooting
• Increase project visibility through traceability
• Manage projects effectively through open and standards-based platforms
• Design, Evaluate and Analyse automation scripts and systems
• Ensure critical resolution of system issues via cloud security solutions
• Improve product quality and help reduce development costs
• Collaborate and coordinate to obtain feedback so as to implement changes

DevOps Engineer Skillset - Engineer Skills required for DevOps

The demand for DevOps engineers is increasing as organizations embrace DevOps. But a good DevOps engineer must possess certain skills, and these can be divided as – a) Technical and b) Non-Technical -

Technical Skills

Scripting Languages – DevOps engineers may be required to provision the infrastructure and hence they will need to know some scripting languages in order to automate it. Some of the more popular scripting languages are – JavaScript, Pearl, Python, Ruby.

Linux Fundamentals – Most organizations have their environment running on Linux. Add to that, Configuration Management tools such as – Puppet, Ansible and Chef also happen to have their master nodes on Linux making it very important for DevOps engineers to have Linux fundamentals.

Grip on DevOps tools – As a DevOps engineer you will be required to have good knowledge and grip on tools used at various stages in the DevOps cycle. These are the tools we had already mentioned under section VI of this article.

Usage of Tools – Knowledge of tools alone is not enough. A good DevOps engineer will know which tool to use at what stage in the DevOps process. In order to facilitate a successful DevOps process, DevOps engineers will need to use relevant tools to ensure continuous integration and continuous delivery.

Cloud and Infrastructure – It would augur well for engineers to have a working understanding of how things work on Cloud and on-premise infrastructure. Infrastructure skills will help engineers develop and deploy applications in an effective manner.

Non-Technical Skills

Collaboration and Coordination – It is but imperative for a DevOps engineer to have good coordination and collaboration skills - after all, the underlying current of a DevOps process is collaboration and unity of teams.

Patience and Flexibility – The likelihood of success is much higher if DevOps engineers exhibit patience and flexibility when it comes to shifting from one area of software development to another.

Effective Communication – This skill goes a long way in helping engineers perfect the art of being a top DevOps professional. Remember, effective communication is what will break siloes between people and teams. Not just that, it will also help engineers develop quality products as they will be able to understand and communicate issues and take feedback in an effective manner.

How can I get started with DevOps?  

Now that we have had a fair amount of understanding on DevOps, it’s benefits and the roles and responsibilities, there is a big question staring at us –

How do we get started?

The Learners’ Journey 

First and foremost, it would augur well to keep in mind what we started off with – that DevOps is so much about culture and collaboration. So, approach the learning with this mindset and you are likely to do well in the long run.

Now coming to how you can get technically better at DevOps –

Read - Spend time and go through a host of material available online. There are several websites and forums that provide material and guidance on DevOps. Start from the basics and yes, it’s easy to get sidetracked with such exhaustive information, so, keep a tab on what you need to start with.

Watch – Thank YouTube. It has some very interesting and relevant videos on DevOps. Watch through these and you will be able to understand the concept better. Browse through any animated series on DevOps basics or fundamentals, they are easy to understand and grasp.  

Test – It is important that you test yourself on your basics. There are some free tests available online, make use of these.

Courses – Finally, the courses. Online learning has reached a new peak. There are plenty of companies teaching DevOps courses with the help of subject matter experts. Research and select the right training partner/Edu-tech company who can help you with not only understanding the DevOps basics but also with how you can chart and build a great career.

Prerequisites – It’s important to also have some prerequisites in order to excel as a DevOps Engineer. Here are some -

• Passion to code, design, and build
• Great interest in automation
• Linux Administration Skills
• Good at one or more of the scripting languages such as Java, Python, etc.
• Knowledge and usage of various tools – Configuration management, Monitoring, CI, CD, etc.
• Understanding of containerization
• Good Troubleshooting Skills
• A lot of patience …

Hands-on experience matters 

Now that you know what would make a good DevOps engineer, let’s look what DevOps ‘’feels’’ like. You can do that by getting into the shoes of a DevOps engineer.

One of the easiest ways to obtain the real-life experience is by interning as a DevOps trainee or engineer. You will know the real picture; you will learn to code, develop, test, deploy and code again. Never lose out on a chance to get a first-hand experience, it will go a long way in helping to do well in your interview and career.  

Remember, when you face an interview or even run into a potential employer at a conference/event, you will be expected to explain about DevOps in a non-technical way. Know that DevOps is first about culture, collaboration and team unity. Keep in mind that DevOps isn’t all or only about automation and technology. Yes, it is also true that tools are used to deliver the ultimate objectives of – better quality product, customer satisfaction and business growth of your organization.

It is likely that you would end up as a good DevOps engineer with focus and hard work on learning the technical aspects and tools but great DevOps engineers and future DevOps leaders will be those who look at the larger picture of what DevOps can achieve for both the organization and its customers. ,

Conclusion 

After all of this, where is DevOps headed? 

Having understood DevOps, we can say with much confidence that this practice is here to stay. Why? Because it does what most technology methodologies do not – breaks siloes and creates unison between the stakeholder teams both technically and culturally. DevOps is a symbiotic practice that gives added advantage and benefits for businesses adopting the practice.

The sheer need to improve application quality, its performance, faster time-to-market and of course, the end-user experience is what will have businesses embracing DevOps in a never-before-like manner.

In more ways than imagined, DevOps is a blessing for businesses out there.  

Internally, it provides businesses with a culture-rich, collaborative and progressive environment. Externally, it provides businesses with the required competitive edge to sustain and move ahead. As the world of business moves into the sea of digitalization and automation, DevOps is that ship helping businesses sail smooth, fast and far.