When people talk about digital transformation, cloud migration is often the center of the discussion. The attention is deserved: organizations that undergo a successful cloud migration can see dramatically improved technical, financial, and operational flexibility and scalability.
The reality, however, is that it’s rarely as simple as moving everything to the cloud to realize these outcomes. Cloud migration isn’t one size fits all. What cloud-forward organizations don’t talk a lot about is that many still rely on data centers to round out their technology operations and serve as the foundation for their digital transformation initiatives.
We’ve seen hybrid environments prove to be an effective strategy for IT leaders who want to achieve dynamic business objectives while minimizing cost and maximizing productivity.
Here are five reasons why data centers play a vital role in enterprise technology and IT innovation.
1: Not All Workflows Are Ready to Migrate to the Cloud
Technology and IT leaders who hear tales of organizations that have moved all of their workloads to the cloud may be tempted to give it a go themselves. The problem with this goal is that, while the prospect of fully cloud-based storage sounds enticing, it’s often not in line with the needs of current company-wide workflows.
Some applications are simply not suitable for the cloud. A number of applications that remain critical to operational success either weren’t built for the cloud or can’t be restructured for the cloud, including:
- Static applications
- Legacy, non-cloud native applications
- Data-intensive applications or usage-intensive datasets
There’s also the likelihood that an organization has some applications that aren’t widely used by many employees. Those applications wouldn’t benefit from the increased scalability and flexibility in the cloud if their usage rates are low and fixed.
2: Non-Elastic Applications Are Better off in the Data Center
Organizations that realize the best returns on investment and the largest impact on their business deploy cloud adoption in three main areas: elastic compute, archival storage, and services. The common denominator in these three types of applications is that native elasticity is written into their architecture.
We're seeing a lot of cloud adoption in the services area, especially because these cloud-native platforms are designed to scale within the cloud. In fact, many of the platforms at the cutting edge of IT innovation have an elastic design that's incompatible with a data center environment, such as:
- Augmented reality and virtual reality
- Internet of Things
- Artificial intelligence and machine learning
- Media services
- Developer tools
But this means that key non-elastic applications, like steady-state computing and production storage, are often better served in a data center. More often than not, moving those assets to the cloud is untenable. If the environments you want to migrate aren’t architected to support cloud, you’ll be tasked with near-impossible refactoring and rewriting projects.
3: Cloud Ingress and Egress Fees Can Be Insidiously Costly
Structural dissonance is one drawback to migrating unsuitable applications to the cloud. High costs are another.
We've seen organizations blindsided by five, six, and seven-figure monthly bills for data ingress and egress in the cloud. These fees can become a significant financial strain practically overnight.
While most cloud providers will be upfront about data ingress fees, egress fees are often hidden in the fine print. When you move data out of the cloud for routine transactions, you might be surprised to discover after the fact that you’ve been charged for every gigabyte you transferred. You’ll also find that static applications typically cost more to move to a public cloud.
When considering a cloud migration of any scale, IT leaders need to know what and how much they’re paying for every application that might live in the cloud. Many may realize that the more cost-effective option is to keep potentially budget-draining applications in the data center.
4: Some Enterprises Lack the Resources to Govern Cloud Platforms
One central question technology and IT leaders must ask themselves before embarking on a cloud migration project is this: do we have the skills to manage the day-to-day governance of a cloud platform?
The truth is that in most cases, organizations don’t have this experience yet – they’re only familiar with on-premise data center management. Even if technology and IT leaders feel confident that their teams are up for the challenge, cloud governance policies are infinitely complex and can wind up stealing much of a team's time and energy.
In many instances, organizations that are working on moving services like IoT and AI to the cloud still need consistent support from the data center from an infrastructure perspective.
5: Cloud Security and Compliance Is Risky for Some Applications
Many organizations may be hesitant about a cloud migration because of strict security and compliance requirements. There’s a lot that can go wrong if the migration of sensitive data sets isn’t executed properly.
This is where a hybrid approach can help IT and technology leaders shore up an overall risk mitigation strategy. While applications with intense security or compliance needs may be better protected in a data center, those that don’t have those characteristics can move to the cloud.
A Hybrid Strategy Brings the Best of Both Environments
There’s no doubt that the cloud has some clear advantages over the data center. But at the same time, the data center continues to provide stability, affordability, and utility to organizations undergoing a complex digital transformation initiative.
The data center thrives as a trusty sidekick to the cloud’s impressive victories. It may not get the appreciation it deserves, but the cloud’s victories wouldn’t be possible without it.