Will data centers will go away as the public cloud grows like crazy? No. Both are growing to keep up with the booming economy. But getting them to work and play well together will be a challenge.
You may also recall the hybrid cloud, which was a paired private cloud and public cloud. Most in the tech field now consider hybrid cloud out of date, and instead push the concept of “hybrid IT.” You can think of that as a private cloud (really, data center) connected with anything that’s in the public cloud.
Today, you need to build fluidity of data sources between systems in the data center and systems in the public cloud. You also need common management and governance tools across the public cloud and on-premises systems. And you need common security, governance, and operations so you can innovate, deploy, and manage on all platforms.
So, what can you do to make this happen? Here are three suggestions:
First, think data
Synergy at the data level is the biggest problem. There is no single source of truth, and databases are decoupled. You’re using 2018 models and databases in the public cloud and 2003 model and databases on premises.
The fix is pretty simple: Get the same database types on both ends, plus middleware links so there is not a demarcation between the data in the cloud and the data in the data center. This typically means you have to modernize the databases on the legacy end, which is a tough and expensive thing to do. However, the alternative is the death of a thousand cuts from the inefficient as-is state.
Second, think common security
This is another tough thing to do, but it’s getting easier. It requires adoption of a security approach and technology that can span both public cloud and on-premises systems.
This typically means adding a common directory between the data center and the cloud, as well as moving to a more sophisticated security system, such as identity and access management. It involves more money and risk, but the alternative is worse.
Third, think process integration
How do you make applications share processing between the data center and the public cloud? It’s a matter of figuring out how applications can run across platforms.
Potential solutions to consider include Kubernetes and containers. However, you also need to understand how to bind processes that exist on both sides to form meta-applications that span on-premises systems and the cloud.
