Related topics
  • ,
  • ,
  • ,

How to introduce modern architectures and best practice to the data centre

Easing the pain

An aerial view of a building in the shape of our vulture logo

We asked three experts, an analyst, a sysadmin, and a vendor consultant, to give us their views on how best to introduce major changes to data centre operations. Their contributions are below.

Tony Lock, programme director, Freeform Dynamics

It is a standing joke in IT that the answer to the question of how to arrive at the promised nirvana is always, “Well I wouldn’t start from here”. But just chucking all legacy systems into the bin and beginning from scratch is rarely an option.

Indeed, last December we asked how IT professionals saw their data centres evolving.

We have posed this question to Register readers regularly over the past few years, and the answers tend to be pretty consistent whatever the context

Unsurprisingly, very few see their organisation undertaking a single transformational initiative to modernise everything across the board. Many more expect to build a modern environment for new projects and leave current systems alone.

At the other end of the spectrum, significant numbers expect nothing much to change for the foreseeable future while modern architectures and tools creep in ad hoc, piece by piece.

The most common approach, cited by more than 40 per cent of enterprise respondents, is for organisations to build a modern environment for their new services and migrate existing systems to it incrementally.

This methodology has proved its worth in many research projects we have undertaken over the years. Establishing a beachhead and then expanding usage over time is recommended by successful early adopters.

The Big Bang approach, in which you migrate everything at once, is too expensive or risky for all but the most confident and financially sound organisations.

When it comes to building the beachhead, it is essential to create a complete solution, with all the elements in place but designed on a small and manageable scale.

This is different from building an environment that just picks off part of the problem – for example advanced server management, ignoring storage and networking.

Some readers suggest pulling the best storage, networking, server and applications people into a virtual team to set things up. One advantage is that from the outset the beachhead is seen to be an integrated initiative, not favouring the interests or priorities of one particular domain.

As everyone starts from different places with different objectives, no methodology for introducing modern architectures and best practices is guaranteed to succeed.

Perhaps the most essential piece of advice is to make sure you have a complete picture of your position.

It is surprising how few organisations can confidently say they possess accurate information about the business services they have deployed, who is using them and for what. Even more surprisingly, many organisations do not even have basic, up-to-date inventory data on their physical systems.

So before you move forward, be sure you know where you are starting from

Trevor Pott, IT consultant

For outside suppliers, the introduction of modern architectures and best practices varies according to the company you are targeting.

Small and medium-sized businesses (SMBs) with limited available capital and manpower are rarely able to implement best practices. Larger organisations, which often need to be compliant with various regulatory schemes, have a clear business requirement.

Larger organisation can follow a more traditional approach. They almost always push technology and practices down from the top, procedure is everything and change management can often be its own department.

Your target is the chief information officer (CIO). He or she needs to be taken to the swank hotel, expensively wined and dined, and coaxed into believing that your vision of the world is the only correct one.

This has been true for ages and it will continue to be so until the current crop of MBAs shuffle off and are replaced by CIOs raised from the ranks of actual IT practitioners rather than business, sales or marketing.

You need only pull the linchpin and these types of executives buy more on feelings of camaraderie and personal trust than on the technical strength of your solution.

By the time you have made it to the top you are cynical, jaded and suspicious of change

Pitching to a proper CIO who has built a career at the coalface is completely different. If you have come up from the ranks, then by the time you have made it to the top you are cynical, jaded and suspicious of change.

Experienced tech veterans will have seen it all go horribly pear-shaped more than once, usually because some yahoo thought that "new" automatically meant "better".

The idea must be sold on merit. What is the return on investment of this latest trend? What is the total cost of ownership? What are the goals of the wondrous new thing and can you achieve the same ends with minor tweaks to existing technologies and procedures?

To convince these people to ride the next wave, you must convince them that their current deployments are inadequate. That means providing a case-by-case analysis and perhaps overcoming a degree of pride in systems that the CIO may have helped build.

Your only really safe bet is a combination of targeted content marketing and hands-on demo time. Get the gear into their hands. Stage demos, trials, virtual labs, whatever it takes. Get the CIO in front of the interface and show what your widget does.

Pushing change from the bottom of the organisation upwards has a better chance of success. Internal folk know the ropes, they know the political situation within the company and they know where the bodies are hidden in legacy technology deployments.

For this reason, it is often better for external vendors to target the lower ranking sysadmins. These sysadmins can put together demos that highlight specific pain points within the company – the only true way to sell change to the experienced CIO.

Small businesses without a CIO are another matter entirely. These are often companies with no change management, paltry resources and an infrastructure that is a series of band-aids applied one on top of the other. They are not strategic planners, they are firefighters.

Sysadmins in these companies are rarely oblivious to the flaws in their network design. They are all too painfully aware – but constrained by resource availability. They solve the largest and most pressing problems while relying mostly on the "too small to notice" principle to avoid more subtle risks.

You will find SMB sysadmins ready to block-replace everything they have, if a means can be found to pay for it. What is in place is crap, they know it is crap, but it is the best crap they could afford; it was build one fire at a time, not as a planned block.

In this case, wrapping your solution up with financing options – or offering the ability to use a lot of what is already in place but in a new way – will win out.

Angus Foreman, chief technical officer, Microsoft Consulting Services UK

First, an obvious statement: there is no single right way to introduce new technology architectures. It depends on the company, its goals, its appetite for risk and its ability to align its business and IT goals.

Organisations with new business models can adopt brand new architectures. Netflix's video-on-demand service is a much discussed example of a model predicated on cloud-based services.

Similarly, Microsoft Consulting in the UK has enterprise customers to which it is delivering complete new architectures. For some customers this means the latest suite of business productivity services (mail, collaboration, presence, for example).

For others it is developing solutions to support new business processes, building on top of cloud-based services.

Often these solutions are storing very large amounts of data to generate customer services that would have been very expensive and complex to develop on premise.

Anecdotally, we are finding that the organisations that are able to exploit Big Bang architecture changes like this have aligned their business and IT goals. They understand the trade-off between benefits, risk and compromises and this enables them to make bigger changes.

As a practical example of such a trade-off, end-users might have historic expectations of the functionality they want from a mail service, such as how much mail storage they have and how they categorise and archive mail using organisational practices developed over many years.

A latest generation cloud-based mail service will provide many new end-user capabilities along with IT benefits, such as more per-user mail capacity or a set of archiving options which might not replicate the organisations customised implementation.

The organisations that are able to align end-users’ goals with the IT goals and the business benefits of a cloud-based service usually make the most innovative changes in architecture.

Can the organisation's end-users be helped to accept the trade-off?

For example, can the organisation's end-users be helped to accept the trade-off between adopting a new archiving model in return for the advantages in mobile and web access and information management capabilities of a latest-generation mail service?

Applying technology principles such as "configuration before customisation" can help with such alignment, and this is something we are increasingly spending more time discussing with our customers.

Organisations that have made significant IT investments must make a more pragmatic set of choices, but new architecture are introducing models that offer opportunities for loose coupling of architectures.

For example Microsoft SharePoint 2013 provides the ability to build SharePoint apps that are surfaced when a user interacts with SharePoint, but the apps are built and hosted independently of the SharePoint platform.

They might be hosted on a wholly separate Microsoft Azure cloud service or in a Linux environment, and can be written using a model familiar to web developers such as ASP.Net or Ruby on Rails.

Such applications obviously have to conform to an API but have no deployment footprint on a SharePoint server, which has benefits both for on-premise and cloud-based SharePoint implementations.

Such an approach allows an organisation to exploit the SharePoint platform without introducing a complex set of integration or customisation issues.

These two examples offer an insight into how Microsoft Consulting is helping customers innovate on modern architectures. ®

Sponsored: Designing and building an open ITOA architecture