Gartner analyst Craig Lowery once described “Hybrid IT” as “connecting, managing and getting the full value out of all the infrastructure and services available wherever they are, including on premise, in the cloud and at the Edge”. Often clients will refer to Hybrid IT as a combination of on-premise based assets interacting with, or even integrated with, cloud-based assets. This is often a byproduct of partial migration to the cloud, where not everything can be placed in the cloud for technical, organizational, or economic reasons. In some cases this is an interim situation, and in others, may be preferred by design. But a key question arises – In an Intentional Hybrid IT design, what goes where, when, why, and how?
While many Hybrid IT instances may have begun with attempts to integrate Cloud -based assets with those not ready for migration, we are seeing much more in the way of Intentional design, where the non-cloud-based assets to be integrated are where they are for explicit reasons, not just accidents of timing. Intentional design then, would balance overall application attributes such as business criticality and integration level required , with application requirements such as for agility and scalability and innovation, and lastly, technical specifics, such as the considerations for low latency, data volume generated, or network dependency. The difficult task then, has been to proactively examine workload characteristics and determine in a consistent and explainable fashion why portions of the application are placed in the array of locations and technologies. What is needed is an easy-to-use tool.
In recently published research titled Tool: Workload Placement in Hybrid IT, analyst Henrique Cecci has developed and presented a comprehensive tool that prompts the user for workload and organizational characteristics, and then provides detailed assessment of where the workload should reside. It uses well proven Gartner models as an underpinning, and when tested, it placed workloads for edge computing implementations exactly where the traditional manual process would have, with the benefit of less work, greater speed, and well documented rationales. I would urge those planners and implementers faced with Hybrid workload placement tasks to give it a try.