Why Software Defined Data Centers and DevOps make IT truly Agile
In this day and age, the business wants to serve its customers quickly and exceed expectations, and it uses IT to achieve this aim. For an organization – and particularly its IT department – this means having the capacity to rapidly adopt innovations involving technology and/or processes. A ‘conventional’ IT department is often unable to fulfill wishes of this kind from the business within the current structure of its technology, organization and processes. And definitely not at the high speed that the business and end customers require. IT departments have long been looking for ways to resolve this dilemma. This time, help is available from various quarters: technology offers the software-defined data center (SDDC) and, increasingly more often, process and development are using various forms of DevOps. Together, these developments ensure that the old need for fast and agile IT can be fulfilled!
This white paper provides all stakeholders – IT and business managers, and members of staff – with insights into how a set of new technologies and modern methods can lead to an IT environment that is optimally agile and fast, and of high value for the business.
Sustainability integrated within the Value Chainleft
Software Defined Datacenter: everything is Software
Many organizations are reconsidering and restructuring their IT landscape. Digitalization is making it necessary to become a digital enterprise in which the business and IT merge into one. Business and technological developments therefore also play a major role in the restructuring of the IT function. What needs to be taken into account when making decisions concerning the retained IT organization (the new IT department)? Traditionally, IT infrastructure is something that The quest for optimal IT flow once built, does its work for many years while remaining unchanged. You could compare it, for example, to the infrastructure of the road network: it is added to or expanded, but organically and not very quickly. The same applies to IT infrastructure. Once built, it is expensive to implement changes and users are stuck with it for a considerable length of time. Changes usually go hand in hand with high costs and related investment considerations.
It is, however, important for the business that IT infrastructure becomes more flexible. Flexibility is needed to provide the business with support quickly in the areas of software development, data analysis and innovation initiatives. And such flexibility is traditionally sought by adding more hardware which significantly increases the complexity of the IT infrastructure as a whole. After all, more hardware means more configuration and more points of failure. IT departments are aware of this and they are looking for different ways to keep pace with the needs of the business. The solution virtually every IT department will choose – or has already chosen – is to move to a form of strategic IT sourcing. The total package of IT services is then made up of a combination of cloud sourcing, third-party data centers and the solutions purchased by the business. In most cases, this ‘modern-traditional’ way of expanding capacity enables IT to respond more flexibly to the needs of the business. However, the required speed is lacking.
DevOps provides the platform for survival
In order to win in today’s fast paced world, companies must fundamentally change the way they build and deliver software. Especially applications exist in an environment of dynamic business needs and at the same time need to remain stable and reliable, whilst depending on constantly changing environments that support them. The ability to roll out business capabilities continuously will be the difference between companies that can evolve and ones that stagnate. This is the background that is driving the adoption of DevOps. DevOps (an amalgamation of DEVelopment and OPerationS), like most new approaches, is only a buzzword for many people. In broad terms, DevOps is an approach based on lean and agile principles in which business owners collaborate with the IT development, operations and quality assurance functions to deliver services in a continuous manner. Thus enabling the business to more quickly seize market opportunities and reduce the time to include customer feedback into business products and services. DevOps leverages the interdependence of IT development and IT operations to create higher quality products and services. It aims to help an organization rapidly deliver the value sought by the customers of IT.
DevOps refers to an organizational philosophy and the professional movement that advocates communication and a collaborative way of working between IT development and IT operations resulting in the fast flow of planned work while simultaneously increasing reliability, stability and resilience of the production environment. Every business wants to move faster without compromising the reliability, stability and security of their systems. And these two seemingly opposing sources of agility and reliability are now possible for every organization to achieve thanks to maturity of DevOps practices and tools.
The 2015 State of DevOps report states that high performing organizations are more agile. They deploy code 30 times more frequently and deploy 200 times faster than their peers. Furthermore, high performing organizations are also more reliable: they have 60 times fewer failures and they are able to restore service 168 times faster than their peers. From the business perspective, we also know that strong IT performance is atrue competitive advantage. Companies with high performing IT make more money and happier employees who consequently, as the research shows, tend to do better work. DevOps promises a great deal in terms of improvement when compared to traditional IT organizations. However, the DevOps philosophy needs to take root in exactly such an environment. Typically, we see (among others) the issues below in traditional IT organizations:
- Inefficient processes, low productivity
- High and unclear costs, ballooning projects
- Weak connection with the customer
- Poor time to market and poor service quality
- Lack of scalable solutions
- Teams working in silo’s
The key problem is that moving to DevOps is a leap of faith, increasingly supported by success
What is the Software Defined Data Center?
The software-defined data center (SDDC) is a radical paradigm shift that transforms the traditionally infrastructure-centric data center, with its focus on ensuring the proper operation of computing, networking and storage elements, into an application or business-service focused environment. The software-defined data center revolves purely around application workload demands, allowing business users to deploy and run their applications in the most efficient and SLA-compliant manner.
In order to install and integrate new hardware, all parties need to perform a great many actions in series and, for the most part, manually. This technology can make the path wider or narrower but will not by definition make it much faster. However, the current state of affairs of technology actually makes it possible to increase the speed of the path, either in steps or all at once. Today, all manual actions for allocating server or storage capacity can be performed in a fully automated or robotized manner. Infrastructure of this kind, known as converged/hyper-converged or software-defined infrastructure, ensures that hardware and software are unlinked. Ultimately, all the functionalities of a data center – networking, computing, storage and security – can be managed as if they are software. Software-defined data centers make it possible to widen or narrow the path in a surprisingly short space of time. For example, whereas in the past it took weeks to create a new testing environment, it can now be created in minutes.
Success factor 1: DevOps is about creating flow.
DevOps taken to its logical conclusion is an exercise in creating flow. The Lean principle of flow means to ensure that all added value work is carried out without waiting time between the process steps. Thus ensuring that customer value is delivered in the shortest possible time. DevOps teams must focus on the flow of work. The challenge is that there are two directions of flow.
Investing in flexibility quickly pays for itself
Shifting to software-defined infrastructure requires a considerable one-time investment. The existing infrastructure usually consists of virtualized (Windows and Unix) blade servers which have been purchased or leased, or are in an outsourced environment. In all cases, heavy investment is involved and/or contracts with an expiration date. Switching all at once thus means incurring additional costs because writing-off has to be accelerated or contracts have to be bought out. Generally, within traditional environments, there is a large group of technical administrators who are responsible for the configuration of the entire server park. The specializations include server administration, network administration, storage, etc. This group of specialists aligns the various technical components with one another and constantly fine-tunes them. When the shift is made to software-defined infrastructure, a large proportion of these specializations are no longer necessary because that work is now software-based. These staff members therefore become redundant for the organization regardless of whether the organization remains in-house or is outsourced. After all, suppliers also use software-defined environments and can therefore not offer these specialists a place. A third point is that a software-defined environment is currently significantly more expensive to purchase than blade server hardware.
The shift to a software-defined infrastructure environment requires heavy investment in hardware, software and people. This may or may not be an obstacle in shifting to software-defined infrastructure, depending on how the organization deals with one-time costs in the business case.
Practical experiences, however, show that such investment is very worthwhile. Significant cost savings in the technical administration of the infrastructure – amounting to between 50% and 70% – can in fact be realized. This is mainly because a converged environment consisting of a single whole (computing, storage and networking) is created which allows for efficient and effective downscaling and upscaling. With an average payback term of about 1.5 years, the investment pays for itself within a limited time. With a software-defined data center of this kind, an infrastructure can be built that is in line with the requirements and wishes of the business and can also keep up with new technologies. As soon as an innovation proves successful, this flexibility enables fast upscaling and, when necessary, the appropriate downscaling. With the SDDC, a highway is constructed that can ultimately be used to facilitate the speed of the business to the optimum. It is self-evident that an IT department can increase its business value enormously through a short time-to-market and by adapting its scope to suit demand. A beneficial side-effect is the ability to deliver a more stable environment, since building a new infrastructure in the event of failure takes a matter of minutes rather than days. The result: improved business continuity and better availability.
As a general rule, the wishes and requirements of customers are becoming increasingly more demanding and must be realized quickly or the competition will take over. Modern development methods such as DevOps are being used more and more often to satisfy customers. The focus of these development methods is on the fast delivery of components so that they can be used quickly. To put things in the right perspective, let’s return to the origins of DevOps. The term DevOps emerged in 2009, in Belgium during the so-called DevOpsDays. The aim of that conference was to bring together IT experts from the development side as well as the operations side of organizations.
And this actually places the term DevOps in its context: a multidisciplinary team that has full responsibility for the continuous operation and development of a service. Examples of companies that use DevOps in combination with continuous delivery are Amazon, Google and Spotify all of which release dozens of changes on a daily basis. Wikipedia has an interesting definition of DevOps: DevOps is a software development method that stresses communication, collaboration and integration between software developers and information technology (IT) operations professionals. DevOps is a response to the interdependence of software development and IT operations. It aims to help an organization rapidly deliver the value sought by the customers of IT. The key point according to the above entry is that DevOps is about producing software quickly. The definition neglects to mention the true value and intention of DevOps. What is the real value that can be realized through DevOps if the approach is used to its full extent?
DevOps and Agile software development
This approach does not build software using the traditional waterfall method (delivering as much functionality at one time for a project) something that requires months of preparation. Instead, software development increasingly takes place in a manner that is known as Agile/Scrum. This method entails the repeated delivery of small pieces of software that are immediately available to the business. Because the pieces are small, the entire process of designing, building, testing, acceptance and deployment can be completed quickly. Moreover, the quality of the software is significantly higher. It is important to take all operations aspects into account from the outset. This can be done by making IT operators part of the development team, or by making the team itself responsible for the operations aspects of its own software. This approach is known as DevOps: a combination of the words Development and Operations. DevOps is of great importance in realizing an environment in which high-quality software can be deployed quickly and in a reliable way.
Success factor 4: Demonstrating Lean Leadership behavior
This is where we come to the role of leadership. One of the benefits advertised about DevOps is the self-organizing character of the multi-disciplinary teams. The automatic reaction is that managers see straightaway see a cost-saving: no need for the team management layer!
The real value of DevOps
In the final analysis, DevOps is about delivering a permanent IT service for a customer and adding maximum value for that customer. This is completely in line with the recent trend which shows that IT organizations are focusing more on the needs of their customers. Information management was a start, the introduction of agile-like methods is a second step. But these methods are sub-solutions to a fundamental problem within IT: the retention of a traditional organizational form in which disciplines are organized separately. DevOps breaks down this traditional organizational structure of IT departments. This makes the approach a highly desirable development with the potential to change the world of IT dramatically. DevOps promises to add value to the operational processes of the business. To capitalize on this potential, IT managers need to be prepared to make their current organizational forms subject to discussion, even if that means their own ‘empire’ will become smaller. In practice, DevOps is regularly linked to Agile/ Scrum in an integrated way. It is certainly true that these methods provide a DevOps team with a number of useful insights. Time-boxing and using a product backlog can be especially helpful. However, these methods are not essential prerequisites for the functioning of a DevOps team. DevOps offers the opportunity to swap a bike for a Tesla. A car to which extra functionalities can be added on a daily basis by downloading new software. This means it is no longer necessary to take the car to a garage or, in the worst-case scenario, to have to buy a new car.
The impact of Software Defined Datacenters on IT
Implementing the software defined data center and DevOps has the potential to be a real game changer for IT. The changes needed to realize this are not only technical, process-based and organizational in nature but are also related to IT sourcing. In our practice, we have identified the following major changes:
The impact of softwaredefined data centers on IT
Implementing the software-defined data center and DevOps has the potential to be a real game changer for IT. The changes needed to realize this are not only technical, process-based and organizational in nature but are also related to IT sourcing. In our practice, we have identified the following major changes:
Integration of IT components
The business has the justified expectation that all separate IT components work together as a single organism. This implies that the various IT services are orchestrated or directed, which requires in-depth knowledge of the different components.
In-depth knowledge of developments among cloud services providers
It is vital for IT departments that make use of IaaS solutions to have in-depth knowledge of developments involving cloud services. For example: Two years ago, Amazon had a limited set of some 10 IaaS solutions. These days, there are a large number of providers and many, slightly different solutions. Which solution is best suited to your own situation?
Making the retained organization flexible
In virtually all cases, part of the IT service provision remains in-house. There are many good and varied reasons for this which involve compliance, risks, privacy and security, as well as critical differentiating processes and/or data depending on the industry concerned and the type of products/services delivered by the organization. In addition, as far as distinctive systems are concerned, the development of new applications and the modification of existing ones often still takes place in-house. Due to outdated architecture and infrastructure, the retained IT organization frequently has difficulty in keeping pace with new technology in terms of performance and availability. It is then especially hard for the IT department to deliver business value quickly. The department has to learn to keep up with developments on the market, and the IT chain is only as strong as its weakest link.
Coordinating the different rhythms
After a software-defined data center has been set up, the IT environment consists of two main strands: the existing solid environment and the new, flexible and agile environment. Each strand has its own rhythm, its own patterns and features. This is known as bi-modal IT. So, for a few years, the IT department faces the challenge of following, managing and orchestrating a twopronged strategy.
The IT organization as we know it has come to an end. The IT department has to find its feet in a new role. It has to move away from being an initiator of IT facilities and a technical administrator, and become a facilitator and director. This is a step that brings the IT department closer to the business.
Facilitating sustainable employment
The question is rightly put as to what exactly is left over for an IT department. The impact is great in terms of staffing. Recent cases show that within Infrastructure departments a 50-percent (or higher) reduction in staff is not at all unusual. In addition, the nature of the work of the remaining personnel and their activities will be significantly changed. To handle this properly, it is vital to deal explicitly with the personal development of the staff involved. Staff members need to be aware of their current competencies, of the new competencies required, and of their personal ambitions in this regard. Employers must give their staff the opportunity to develop in accordance with the new reality.
Using the SDDC in combination with DevOps does not in itself give rise to significant new sourcing challenges. The development of DevOps and a SDDC within an organization frequently does not stand alone. Often, there are ongoing SaaS initiatives running within various parts of the organization aimed at fulfilling the needs of the business. This requires governance of the SaaS services provided and orchestration of the SaaS solutions. The result is that the IT staff needed must meet a different profile.
How do you set the transformation in motion?
It is desirable to have the implementation of the software-defined data center and the DevOps approach run simultaneously. After all, they reinforce one another. If one of the two lags behind there is a risk that no added value will be realized. The weakest link determines the strength of the chain. Investment is required when setting up a software-defined data center but this investment will quickly be translated into business value.
It is not advisable to have the transformation take place like a big bang change. In practice, it is better to have the implementation take place incrementally. In the new world, upscaling is simple. This means that a big bang scenario is not necessary. Moreover, it is unappealing from a financial perspective due to the additional necessity to write off the existing infrastructure. There is most often a logical implementation moment for each component of the IT infrastructure. Such moments are generally linked to the lifecycle of the infrastructure and/or an application domain. This means that the new world can be set up piece by piece. An additional advantage of the software-defined data center is the possibility to run a trial set-up for a time and thus acquire insight into the user experience. This information can be used to lay down a roadmap. When setting priorities, the applications with a high change level are the first to be migrated to the new world. The business value of the software-defined data center is highest for these applications. The implementation of DevOps has similar characteristics. Here too, it is sensible to avoid a big bang implementation. Instead, start with a specific application domain and try to learn from the experience before transforming the next application domain. In addition, it is important to realize a balanced change in which the staff, processes and organization involved move forward together. It also applies here that the weakest link determines the strength of the chain ( = improved performance = ).
DevOps teams can use ‘selfies’ for measurement purposes and to create transparency regarding where things are going well and where improvement is necessary. Desired improvements can be prioritized and implemented within predefined time boxes. It is also important here that the DevOps team steers its own improvement and acquires clear insight.
- Software-defined data centers and DevOps together have the potential to be a real game changer for IT and the business. What is involved is the creation of an optimally agile IT environment against lower costs.
- Software-defined data centers provide an infrastructure that can easily keep pace with the requirements and wishes of the business and keep up with new technologies.
- DevOps makes it possible to provide a permanent IT service and to deliver maximum added value to the customer.
- The changes needed to realize this are technical, process-based and organizational in nature.
- The implementation of software-defined data centers and DevOps takes place in phases. The motto here is: Think big, act small!
- The transformation involves a balanced change in the processes (flow), the organization, the performance of the team, the automation of IT activities using SDDC, and the behavior of the members of the DevOps team.