A decade ago, when I was an industry analyst covering cloud computing and services, one of the compelling things about the cloud’s emergence—in addition to cost and flexibility benefits—was its potential to democratize IT. Developers and engineers would be free from the iron grip of centralized IT that tightly controlled all infrastructure and operations. Ten years later, the public cloud is ubiquitous and saying every company is a software company is a cliché. But internal data centers didn’t go away completely and neither did centralized infrastructure and operations teams. Together, cloud’s universality and the persistence of on-premises data centers have resulted in enterprises adopting a hybrid cloud strategy whether they intended to or not. To adapt, they’ve evolved two distinct IT models: centralized and decentralized IT.
Managing hybrid cloud environments is challenging
Any company adopting a hybrid/multicloud is facing a complex organizational challenge. The idea of the cloud is to give more control to the users of hybrid and public cloud services (depicted in the image below, second row) rather than having centralized IT control.
But with that new power comes greater responsibility. The role of IT in a hybrid and multicloud world is to empower line of business (LOB) leaders (depicted in the top row) to make smarter purchasing decisions, providing them with visibility into how their users are using cloud resources. LOB leaders also need to collaborate and communicate more effectively so that they, and the organization as a whole, optimize their use of cloud services and prevent bad things from happening.
At the same time, IT still has its responsibility to manage On-premises data center resources. Consequently, most enterprises have a hybrid IT model to go along with their hybrid cloud.
On the left, on-premises data center resources are the domain of infrastructure and operations teams that deliver and control it all. Those teams, which include infrastructure and operations staff, capacity planners, and systems/virtualization (VI) administrators, require tools for configuration, provisioning, automation, capacity planning, and security for all data center assets. The centralized model is also a primarily capex model so, while it may be underutilized at times, it’s at least predictable. On the right is the public and multicloud world where cloud architects, cloud operations teams, financial/business analysts, and cloud security teams need a different set of tools. Rather than being capex-intensive and hands-on with operating infrastructure, the decentralized IT model is opex-intensive and requires management software designed for the variable consumption of “aaS” (SaaS, PaaS, IaaS) solutions by diverse teams. These teams need tools to govern usage, optimize costs, provide showback/chargeback, and mitigate cloud security threats and vulnerabilities. Cloud management for decentralized IT needs to facilitate coordination among IT, cloud users, and Finance teams so they can manage opex the way they used to manage capex.
Gartner Research refers to these models as IT choosing to be “in the way” or “on the side.” As an enterprise adopts more cloud services, IT can decide to be “in the way” and force all provisioning of cloud services through central IT. This has the advantage of locking things down, but it limits choice and speed. Plus, with the number of changes public cloud service providers can roll out in a year (AWS pushes more than 2,000 updates per year), IT simply can’t keep up. Alternatively, IT can choose to be “on the side” and provide users with the ability to provision their own resources but provide some framework for governing and reporting on usage. This is exactly what CloudHealth does, providing cloud management that gives users what they need but facilitates an enterprise’s “on the side” model.
Hybrid cloud adds more complexity because it bridges these two worlds. A great example depicted in the graphic is VMware Cloud on AWS. An organization with a sizable data center footprint that is migrating workloads to VMware Cloud on AWS and/or to public clouds needs to have both data center and cloud skills and a way for centralized and decentralized IT to collaborate and communicate.
Hybrid cloud management bridges centralized and decentralized IT
CloudHealth Hybrid, generally available at the end of this month, brings together these two worlds. Cloud operations teams can incorporate VMware Cloud on AWS, public cloud, and vSphere cost and usage data in one management platform. This gives multiple teams across the organization detailed cost and usage reporting, including showback, for all cloud resources. Users also gain a deep understanding of actual VMware Cloud and data center cost drivers and can compare them to industry benchmarks so they can optimize spend across their environment. When an organization combines CloudHealth with VMware vRealize Suite, they have a complete, end-to-end cloud management solution for both centralized and decentralized IT.
Very few enterprises are strictly “in the way” or “on the side”. Most will have a hybrid approach that includes both to reflect their hybrid environment. For many reasons, including things such as compliance, security, or criticality, IT will exert centralized control of infrastructure or service provisioning whether in the cloud or in the data center. However, for the bulk of cloud applications and workloads—whether public or private/hybrid—IT can be more democratic and give users more autonomy by providing guardrails to govern usage. Regardless of how democratic your IT model is, CloudHealth and CloudHealth Hybrid provides all the different constituents with the management capabilities they need to move fast, work together, and make sure nothing is breaking.
