For CIOs a common mantra is:  I need to serve the business with a cloud-like application experience.

On the surface the requirement is simple.

Below that statement there exists a stack of hidden complexities well understood by CIOs. However, press and analyst reports suggest CIO’s are being urged to move everything to the cloud by less informed CEOs and CFOs.

This blog is not an attempt at generalization of all of the complications faced by today’s CIOs. Instead it explores a key layer in the IT stack with an emphasis on management of the physical infrastructure and operating it as a cloud-like service to facilitate flexibly provisioning applications.

 

The best execution of a multi-platform strategy to deliver the cloud experience journey begins among the physical assets.”

 

It is about avoiding ‘saying cloud and you mean data center’ and saying ‘data center when you mean cloud.’ 

All CIOs know the IT stack has many layers. At the base level these layers consist of physical infrastructure comprising mechanical, electrical and IT hardware. Within a modern data center the management of these facility and IT assets has the potential to become more rather than less complex. 

 

Traditional IT Stack

Figure 1. ServerFarm’s unique capability is to take all physical assets in IT, facility and DC environments and return them as a virtualized service. By combining software, people and processes, ServerFarm’s InCommand™ can reveal a unique view of infrastructure.

 

Part of the solution which will enable CIOs to achieve their application delivery objectives, is to take away the pain of day-to-day management of these assets by turning it into a cloud-like service. That service should make infrastructure provisioning and availability to the business indistinguishable from that of any cloud provider. (Cloud-like data center management-A new normal)

No longer just build or buy

Unless you are a CIO whose organization is young enough to have been ‘born on the cloud’ or brave enough to have migrated everything to a public cloud, something that has stayed constant for those with any kind of IT or data center heritage is the requirement for physical assets to be managed.

If we begin the cloud journey with an eye on the destination – say five or ten years down the road – we can ask, what does enterprise workload distribution look like? What we know is change is accelerating and workloads will move around, thus they will need to be flexible.

So, amid these different pressures and constraints the question becomes how best to manage physical assets in order to stay laser focused on provisioning that cloud-like application experience to the business.

 

At an application level the questions include: Where is the best place for me to deploy an application? What is the best execution venue for the workload and what does best execution mean?”

 

Ongoing industry trends continue to see enterprises wish to move beyond ownership of data centers and the associated management responsibility of everything from utility power to wide area network connectivity.

Where once the physical layer supporting workloads and applications was readily identifiable, (I can see it, I can touch it) now it is spread across different assets, platforms and services.

A CIO’s scope of activities is likely to include data center ownership, renting multi-tenant commercial data center space, physical facility and IT physical asset management within dedicated white space or in racks housing physical IT assets, shared or dedicated power and cooling physical assets, managed services for facility and IT assets, private cloud operations, application services from all the big public clouds, renting compute cycles, storage and networking for DevOps, subscribing to serverless computing, infrastructure, platform and software as a service…the list could probably write itself and could go on and on.

Such a broad scope requires the execution of a multi-platform strategy built on a cloud-like service for physical asset management.

Where do we start on the cloud journey?

At an application level the questions include: Where is the best place for me to deploy an application? What is the best execution venue for the workload and what does best execution mean? How to define best? Is it the place that offers the greatest flexibility? The quickest to start? The easiest to deploy? The most secure place to deploy? Or the cheapest to deploy?

 

ServerFarm believes cloud-like application delivery starts with physical assets that can be deployed and behave like a cloud.”

 

It could be some or all of those things depending on the desired outcome.

Within a multi-cloud, multi-platform approach there exist several categories. The first category is cloud. There will be workloads in the public cloud, whether single sourced on AWS, Google Compute Platform or MS Azure or in multiple public clouds.

The public cloud could be defined as a place where the only thing under the CIO’s control is the use of the application and the data. [Link to Exec Opinion 2 Customers want flexible clouds and commercial data centers]

There will be an increasing number of workloads running across multiple clouds in different cloud types categorized as public, private or hybrid. Many new workloads are being deployed on a cloud-first or cloud-only basis unless there is a specific requirement for them to be somewhere else. On a public cloud a CIO doesn’t own infrastructure and nor should a CIO need to worry about any infrastructure operation (except in cases such as compliance on data sovereignty, GDPR or other regulations.)

What does the somewhere else look like?

Other categories of best execution venue would be in your own building, in an enterprise owned data center or in someone else’s multi-tenant data center (MTDC).

As the value of data increases (data is the new oil) this local management and deployment need is set to continue for the next ten years at least. There will be some workloads that have to be maintained locally in owned or leased infrastructure. The criteria for these will include location, performance, resilience, latency, regulatory demands and legacy (there could be old or specialized applications that can’t be cloudified).

 

ServerFarm Oak Brook data center - Outside

Figure 2. ServerFarm believes cloud-like application delivery starts with physical assets that can be deployed and behave like a cloud.

Cloud-like physical stack

In all environments outside the public cloud there is managed infrastructure to be considered.

ServerFarm believes cloud-like application delivery starts with physical assets that can be deployed and behave like a cloud. That in turn means having a service-led approach which allows assets to be managed more effectively.

For the first time the capability now exists to manage and provision physical data center and IT assets as a cloud-like service.

The best execution of a multi-platform strategy to deliver the cloud experience journey begins among the physical assets. For any CIO executing a multi-cloud and multi-platform strategy it is the sensible place to start.