Hybrid cloud is ushering a major change in the role of central IT. Instead of being the infrastructure gatekeeper for the enterprise, it’s becoming more of an IT safe keeper and facilitator. When you look at the evolution of IT environments, this change seems inevitable. Let’s see why.
In the not-too-distant past, the way every user constituency gained access to IT infrastructure capacity was make a request (submit a ticket) to the central IT services team. Presiding over data centers filled with physical and virtual servers (shared infrastructure capacity), this group ensured sufficient infrastructure capacity would be available, workloads would run securely on the proper infrastructure tier, costs associated with deploying and managing workloads were tracked and allocated to the appropriate business unit (or department or team), and so forth. To a very significant extent, we still see this model today. Users submit tickets to request infrastructure, and deployed applications are subject to policies separately reviewed and confirmed by central IT. In recent years, many enterprises have leveraged some form of automation in the service delivery area, but the fundamental mechanics of how a user constituency requests and receives their application via the centralized IT group has really not evolved.
From central IT’s perspective, the control they wield under this model is a means to an end. If they were to give users direct access to the underlying infrastructure, the risks of capacity, budgetary, security or other policy constraints being broken are simply too great. The unfortunate consequence of this service assurance process is that it requires a lot of overhead…and the incremental time required stifles user productivity.
Then public cloud happened. The advent of public cloud brought businesses the ability to move faster, be more competitive, potentially reduce costs, and introduced a number of amazing technological advancements. It also enabled users to go on their own cloud journeys, outside the purview of central IT. Most enterprises are still in the early stages of going “all-in” on public cloud and still rely heavily on their virtualized (VMware) infrastructure, which generally remains under the domain of central IT. As the number of platforms the business uses grows, however, so does the complexity. If central IT services is not overseeing all of the public cloud deployments for each user constituency, then ensuring adherence to budgetary, security and other policy constraints can become a considerable challenge. In many enterprises, those business units or departments end up taking on greater responsibilities in those areas.
So, what is to become of central IT services amidst this change in infrastructure options for user constituencies? Can the team reel in all of those business units and departments and teams? How much of a change is really needed?
Shared Services vs. No Shared Services
The way IT has been working in the majority of enterprise hybrid cloud environments can be considered a “Shared Services” model. In this model, IT manages and maintains the underlying infrastructure for private and public clouds, and gives access to business units, departments or teams, typically with some form of showback or chargeback to tie to costs to users and fund project expansion. Central IT wants this kind of consolidated infrastructure. Having all cloud accounts under the same management makes it easier to negotiate with the cloud providers, and to implement management tools that give IT more visibility. In short, the “Shared Services” model makes it easier for IT to do their jobs, but it doesn’t always contribute to user productivity.
Public clouds have enabled a “No Shared Services” model, where business units adopt platforms independently, pay for them themselves, and are (presumably) on the hook for any security and compliance requirements. The “No Shared Services” model often creates a difficult situation for everyone. It has business units spending time and resources on things they used to rely on central IT for – maintenance, automation, policies, pre-defined workflows and more. They also need to build their own expertise on security, cost management, networking, storage, etc. Often other lines of business in the organization have similar needs, and have already built systems to support those requirements. It’s challenging to reuse those systems when each unit is operating independently.
Business units, much like their counterparts at central IT, are not trying to make anybody’s life harder. The lengths they’ll go to to avoid the traditional IT model shows how unhappy they are with it, and how much pressure they’re under to deliver faster. In many enterprises, the result is that IT is unable to do their jobs due to the loss of control, and business units struggle to keep up with their new responsibilities.
The Federated IT Model
A new hybrid cloud landscape requires a new IT approach that bridges the gap between business units and central IT. For many enterprises, a “Federated IT” model fits the bill. This model gives teams, departments, and business units a greater level of freedom to source their own resources and build their own guardrails and tools. At the same time, it establishes a set of policies and services set by central IT and mapped to the company’s own hierarchy, which ensures security and compliance standards are maintained and costs are kept under control.
With Federated IT, business units can operate with minimal overhead. They can be much more self-reliant, efficient and productive precisely because they can also rely on services, infrastructure, expertise, and tooling from central IT. Central IT has central visibility and is able to enforce critical policies while giving business units more autonomy to operate and adopt new technologies. Everyone involved gets what they need to do their jobs, without putting budgets, security or compliance at risk.
As hybrid cloud adoption grows, IT leaders will increasingly struggle with enabling Federated IT and balancing the competing needs for flexibility and control because the methods and legacy management solutions that were built for how IT worked in private cloud are at odds with the self-service model the public cloud presents. Just as hybrid cloud is ushering in the need for Federated IT, it’s also accelerating the need for a more evolved “Cloud 2.0” management solution that turns central IT gatekeepers into safe keepers.