0
You may have noticed a tendency for tech vendors to overplay the latest trend: if you’re not on the bandwagon, there’s an implication that you are missing out in a big way.

For most businesses, that’s often how it feels when the inevitable topic of cloud comes up. Everyone is doing it, apparently. Except there’s a catch. Despite surging growth, so far only around 23% of our infrastructure has actually moved – for most organizations often only the low-hanging fruit – mail servers, file/print, web infrastructure: in other words, the less complex, less critical workloads.

Today, the reality is that most organizations are operating a mixture of on-premises and cloud-based tech – better known as a hybrid IT model. In fact, hybrid IT is actually the dominant and growing trend among most players. We commissioned research into this, and discovered that the pace of hybrid adoption among established players is accelerating. Unlike legacy-free start-ups, which often start off cloud native, most organizations are weighed down by systems of record that cannot simply be jettisoned – and nor should they.

Of course cloud is a temptation for legacy-loaded players, under competitive pressure to rapidly launch a new digital service or keep up with a new disruptive competitor.

Too often this approach overlooks the long-term impacts of adding a cloud environment into the mix. These may be to do with technology, for example, to win scalability and latency, or could relate to security and compliance: who exactly is responsible for your data? There could also be an impact on any number of other factors, such as orchestration, management, skills optimization – even the very core strategy on which your business depends.

It’s a complex picture with more IT choices than ever – but it’s not always obvious what the right choices are, nor their implications.

This article asks and answers the most pressing questions about cloud and the new world of hybrid IT. To achieve this, we start with the core question: what workloads are candidates for cloud migration and which ones do experts recommend leaving on-premises – and why?

A word of warning, however: amid all these choices, it is vital to remember that going hybrid is not a destination, and no organization should add increased complexity without good reason. The real question is “what do you want to achieve?”. Some of the more compelling answers may include greater workplace flexibility, better service and price competitiveness, total control over critical workloads, or transformation of business processes.

Any consideration of a hybrid architecture needs to start with these strategic considerations, and Fujitsu can provide insight and advice.

This Q& looks at the 10 questions that customers should ask when considering a hybrid model. When it comes to the actual technology choices, workload requirements on infrastructure are very different and require informed decisions. Our goal is to provide answers to the key questions you should ask to determine your optimal hybrid IT strategy.

1

1
Where should we place which
workload, and why?

This really is the fundamental question and absolutely the right place to start. Some workloads are perfect for cloud migration, others could be disastrous.

A good way to frame your approach is to deploy a two-by-two Boston Matrix – a popular and well-proven business strategy decision tool. The trick to obtain meaningful results with this technique is to get the matrix axes right.

For most organizations, the two most important aspects of hybrid transformation are likely to be the complexity of the workload and the business value it generates.

Deciding what represents business value will vary hugely organization by organization. Remember that it can be tempting to attribute value to a workload that is past its sell-by date but looks too hard to retire.

To recognize complex workloads, look for one or more of these factors:


  • A high level of customization
  • Special security, privacy, and compliance demands
    and where you need to keep full control
  • Where low latency is essential

Now plot your workloads onto the matrix. Complex workloads that do not create business value should be retired. If they create real value for your business, run them on-premises. Less complex workloads that create high business value should run in the cloud, because the cloud stands for agility and cost-effectiveness. Workloads which are less complex and generate lower business value may be outsourced on price and run off-premises in the data center of a hosting provider.

2

2
What’s the best approach to build or
refresh on-premises infrastructure?

With workload analysis in place, you can start to look at the details of how to build the optimal hybrid environment. Workloads that should remain on-premises will be both high-complexity and high-value, and choosing the right partner is key.

Let’s start with the hardware side of things. Your vendor of choice must be able to build complete data centers on your premises. It should provide you with a broad range of infrastructure options (server, storage, network, and software products) to match the performance and scalability requirements of your workloads.

Your vendor of choice must be able to build complete data centers on your premises.

Turning to system management, the vendor should offer intelligent and innovative solutions providing all the functions for fail-safe, flexible, and automated 24/7 system operations.

To maximize your investment, contain costs, and remain compliant, the vendor must be certified by the major software vendors, to manage the complex mix of contracts, agreements and license models.

For services, all products should ideally be supplemented by consistent on-site support services covering all your locations.

The vendor should also be able to offer a complete portfolio of financial solutions enabling the planning, transitioning, acquiring, managing, and retiring of your IT infrastructure.

3

2
How to deploy new infrastructure rapidly?

Any workloads remaining on-premises are, by definition, going to be high stakes. Your job is to minimize or, preferably, eliminate any risk involved in the transformation process.

Unfortunately, risk is off-the-scale in do-it-yourself infrastructure rollouts, where all options and components are chosen and assembled by an in-house IT team. Unsurprisingly, this tactic has been going rapidly out of favor in recent years, as data center availability is make-or-break for organizations in a digital economy already desperately short of the necessary IT skills. This has rendered the build-your-own option both unsustainable, and even potentially career-limiting.

The alternative pre-integrated approach, where all options, system integration and testing are handled in the factory, is the norm. Pre-integrated ‘Integrated Systems’ reduce complexity through eliminating the high-risk trial-and-error process when introducing new infrastructure, while the compatibility of all components is guaranteed. This eliminates the requirement for deep knowledge of all components involved.

Less time is needed for planning while deployment is massively accelerated, shortening time-to-value.

The optimized design of Integrated Systems means system resource utilization is also optimized. This has a positive impact on a whole number of things: data center space requirements, cabling, energy consumption and cooling. Integrated Systems represent an optimum foundation for efficient operations and reduced maintenance efforts. These aspects help reduce CAPEX and OPEX costs and refocus on the important aspects of business.

For customers looking for simplicity, via an all-in-one appliance, another software-defined approach is available which also integrates storage –hyper-converged systems.

To accelerate time to value and mitigate risk, we recommend selecting a vendor who provides with a portfolio of hybrid IT-enabled integrated or hyper-converged systems.

4

4
Who will manage which
infrastructures?

For workloads that will remain on-premises here, you may assume that an in-house team will manage these – but that isn’t always the optimal choice.

When it comes to running your new hybrid environment, one main benefit is the low-management overhead. However, you still have management responsibility for your hybrid environment.

This requires a wide range of specialist skills and includes the abilities to consolidate, standardize, automate and virtualize systems utilizing appropriate cloud services, customer-owned IT and network connectivity – to create and deliver efficient, low-cost, flexible IT around the world. However, most organizations struggle to justify the cost and management overhead of maintaining a suitably qualified team. Instead they focus on managing a single MSP with proven capabilities. Other important criteria are using lean methods to build a road map to optimize IT infrastructure, operations and premises, linked to clearly defined business benefits that can be tracked throughout the life of the contract.

For international or global players, the capability to deliver IT services around the world provides assurance that your operations receive the attention to service quality, consistency, and security they require, wherever they are located.

Ideally, the operational delivery processes of the managed service provider should be ITIL®-compliant, the global data centers designed and operated to the standards of the Uptime Institute, with secure operations conforming to ISO 27001, and meet local standards.

5

5
Who will host my infrastructure?

For less complex or critical workloads that provide only limited business value, the appropriate option is to choose a hosting provider, with the relevant infrastructure managed directly by you or the host.

The hosting provider should incorporate green IT principles in its data center operation to minimize carbon emissions and energy costs. Most importantly, make sure the hosting provider has sufficient security measures to ensure that only personnel with the appropriate credentials have physical access to your data.

6

6
How to find the best cloud
provider for your needs?

High-value, low complexity workloads are ripe for migration to the cloud. This strategy has been part of the winning formula for many organizations for some time now. It continues to underpin many of the world’s leading and most disruptive players. Leading cloud technology providers are building an ever-growing range of capabilities.

‘Multi-cloud’ strategies have become increasingly dominant. Fujitsu’s State of Orchestration survey of IT decision makers reveals that customers are blending a mix of nine separate cloud vendors today: Cloud complexity is growing.

Investigating the service portfolio of cloud service providers is a daunting task. It is more than likely you will find that a single cloud service provider cannot cover every need, so you will end up with more than one cloud service. During this process, it can be invaluable to have someone by your side who has developed strong partnerships with all the leading cloud service providers and can provide unbiased advice about the creation of a personalized, multi-cloud ecosystem to exactly meets your requirements.

Given that hybrid environments are the new reality, that partner should also have clear and demonstrable experience in the worlds of both on-premises IT and multi-cloud IT, with the know-how to manage, balance and orchestrate

7

7
What is the most effective way to
go about cloud migration?

After identifying workloads suited for cloud migration, think about how to do it. This might seem too obvious to ask, but there are complexities you might not have considered, which could have future consequences.

Cloud providers are keen to make migration as simple as possible and often waive any data transfer costs. The same is not true, however, if you move in the opposite direction. Then, costs of transferring data between on-premises, competitor or regional data centers can be unexpectedly and uncomfortably high.

This makes it imperative to understand your options and plan for contingencies before making decisions.

One common migration model is the transfer of data and applications from a local, on-premises data center to the public cloud. However, this could also entail moving data and applications from one cloud platform or provider to another – cloud-to-cloud migration. A third type of migration is to move data or applications from the cloud to a local data center – called reverse cloud migration, or repatriation.

Without proper planning, a migration could negatively affect workload performance and lead to higher IT costs. Your partner should be able to provide a Software Defined Wide Area Network (SD-WAN) to ensure you can take control over WAN performance and costs by creating complete visibility and the ability to automate and simplify services and processes. Your provider should have real-world experience of network management – preferably network ownership - so you guarantee avoiding bandwidth bottlenecks.

These are not hypothetical risks – every month brings new examples of major corporations suffering from cloud under-performance or total outage, and the consequent loss of business, damage to brand and reputation, or even regulatory fines.

8

8
How can you make different workload
placement options easily available?

You are now envisaging how these workloads are going to be used day-to-day – a really important question! But finding the right technical solution to workload placement is, in many ways, putting the cart before the horse. Unless people in your organization can find what they need to get their job done - and can use it without days of training – then the entire transformation exercise will fail.

If, like most organizations today, you use a blend of different workload placement options, make sure to introduce a service catalog management tool to provide a unified portal where users can access IT resources without having to know where they are hosted.

Business users get access through an intuitive self-service portal, and can easily find and consume all services, regardless from where they are delivered. Such a tool enables an organization to introduce new services quickly, keep control of service usage, and report and charge service consumption to business units.

9

9
How does hybrid IT affect my
data protection strategy?

The threat of critical data becoming unavailable – either momentarily or permanently – is one of the greatest dangers hanging over any organization, with costs and implications that range from deeply uncomfortable to terminal. The good news is that hybrid IT increases the options for fail-safe data protection.

The need for data resilience is not a new discovery and many IT organizations have already deployed data protection appliances. Best practice is to have a third copy of your data on another physical site in case the main site fails completely during a disaster.

For all organizations that can’t afford dedicated disaster recovery infrastructure, the cloud offers a viable alternative.

Here, production data is backed up to a data protection appliance. This automatically copies data to a cloud service provider. If the production storage fails, a fast data recovery directly from the appliance can be conducted. Should the whole site fail, the last line of defense is in the cloud. Recovery time from the cloud takes much longer, but recovery is still possible.

Orchestrating and automating data flows across a hybrid architecture requires specific technologies and the specialist people skills to build and manage them. Choosing how to provide these capabilities – either in-house or via a managed service – is a vital decision with consequences in terms of cost, management time and effectiveness.

10

10
Who can support me on my journey
to Hybrid IT, end-to-end?

It’s probably clear that many organizations are unlikely to have the depth of in-house experience needed to assess all options in a hybrid IT transformation. That’s simply a reflection that skills are scarce, and IT departments are under pressure to remain lean.

This expertise also needs to embrace a range of platforms, as there is no single cloud that addresses the range of challenges and aspirations that organizations have. At a time of global IT skills shortages, this is a tough brief, especially as CIOs - understandably - remain wary of being forced to compromise with a one-size-fits-all solution from an integrator tied to a single preferred cloud platform.

While many integrators talk about 'end-to-end' capabilities, few truly deliver on that promise. The complete range of capabilities to transform infrastructure, applications and services in the cloud comprises six headings:


  • Legacy Modernization
  • Multi-Cloud Integration
  • Apps Transformation
  • Cloud Native Development
  • Service Transformation
  • Service Orchestration

Fujitsu is among an elite group of systems integrators.

Fujitsu is among an elite group of systems integrators with the skills and experience to deliver hybrid, multi-cloud, multi-modal IT both reliably and at scale, and to do so on a global basis. With hybrid IT becoming the norm and playing such a critical role in the success of any organization today, make sure its management is in the right hands.