Executive Roundtable: Economics of Hybrid Multi-Cloud Computing Can Be Compelling...But Also Complicated

By Lane F. Cooper, Editorial Director of BizTechReports and Contributing Editor to CIO.com

While technology modernization initiatives are driving most large enterprises to adopt hybrid, multi-cloud computing strategies to optimize agility and performance, the economics of operations can take a lot of work to nail down. This is especially true as organizations come to grips with the dynamic nature of shifting workloads in which mission-critical data and applications flow from one cloud to another -- and, in many cases -- can end up back in on-prem infrastructures.

David Anderson and Monte Bauman from IBM

These were among the conclusions of a CIO.com executive roundtable co-hosted by David Anderson and Monte Bauman from IBM. The event featured senior technology executives from the higher education, financial services, global manufacturing and technology sectors. Below are a few key takeaways from the event:

Reconciling CAPEX and OPEX Environments

Executives from large organizations participating in the roundtable reported moving into a "Cloud-Economics 2.0 Era," which calls for carefully recalibrating infrastructure management practices to ensure cost-effective operations. Methods that made sense -- or at least did not carry consequences -- in on-prem, CAPEX-intensive infrastructures can result in costly outcomes in OPEX-oriented cloud environments. 

For instance, conventional on-prem practices employed when responding to business requirements -- such as pausing a workstream to focus on an urgent project -- must be managed differently in cloud environments. As one executive put it: "In the cloud, it is important to close the door behind you when parking a project because the meter in the cloud continues to run. From a cost perspective, you must manage the cloud more than you ever managed on-prem IT infrastructure." 

That said, all participants enthusiastically accepted the important -- and even growing -- role of cloud-enabled operations, especially when rapidly deploying customer-facing applications.

Managing the swinging pendulum of workloads

Like everything in IT, participants noted an ebb-and-flow rhythm to infrastructure operations. Periods of centralized computing give way to distributed architectures, only to be re-centralized a few years later. The same dynamic, several agreed, is taking place when it comes to current technology modernization initiatives. Participants noted a pendulum-like dynamic in which frenzied efforts to migrate everything to the cloud are re-evaluated for a return to on-prem environments. This, executives noted, is happening for various technical, operational and financial reasons. 

Avoiding Platform Dogma 

The pendulum dynamic is a cautionary metaphor for organizations that dogmatically engage in "platform wars." One executive observed, and others concurred, that it can sometimes seem like C-level executives and board directors engage in a "magazine-headline-based approach" to infrastructure planning. It is not often feasible for large complex organizations to adopt an "all-cloud" infrastructure strategy. And if there is any validity to the pendulum theory, organizations risk getting caught short if "enterprise-wide modernization" is directly synonymous with "cloud migration."

Several noted that there appears to be a highly negative attitude among non-technical executives (and shared by some technology professionals) about the venerable mainframe platform. That is why many senior leaders get caught by surprise, one executive said, when the work-horse capabilities of mainframes cannot be replicated in the cloud.

That said, modernizing high-performance on-prem resources should be much more than replacing hardware. While it may not make sense to migrate all activity to the cloud, abandoning legacy practices -- like unfettered software customization and waterfall-based code development -- is prudent. Sustained success will likely result from integrating cloud-influenced best practices -- including agility, DevOps and extensive application configuration -- into on-prem data centers and mainframe platforms. 

The final area of consensus revolved around the notion that modern hybrid, multi-cloud infrastructures will result in a unique set of technical, operational and financial profiles for every organization. No two enterprises will be alike. For this reason, organizations should develop core competencies around the ability of executives, managers and even individual contributors to develop a deep and ongoing understanding of how constantly shifting heterogeneous infrastructures affect the total cost of ownership (TCO) picture.