Cloud is omnipresent, seemingly everywhere and scalable to meet the end user wherever — but behind the scenes, the technology is laid out in regions.
Cloud regions are the geographic locations where data and resources supporting cloud technology sit. Vendors choose these locations differently, and customers lean toward nearby regions for faster availability. But the decision-making process is layered with regulatory complexity.
"Even in the era of terabit data networks and massive scale-out compute architectures, locality matters," Brian Alletto, a senior architect in West Monroe's technology practice who specializes in the cloud, told CIO Dive via email. "It matters for application performance, data sovereignty, service durability, and other use cases that are best served by multiple hosting locations."
Companies in some industries, such as the financial and public sector, aim to avoid lock-in or single points of failure and pick cloud regions accordingly, a Gartner prediction states. In regions prone to disruption, companies will look to different geographically affiliated clouds to maintain stability, or even rely on legislation and regulation to maintain control.
Vendors build cloud infrastructure in multiple regions for several reasons, according to Bob Moore, principal and partner of cloud engineering at PwC. Motivators include moving services closer to the end customer to reduce latency, resiliency against outages, data sovereignty and government mandates.
For users, region availability frequently comes into play when an outage occurs. In the fall of 2020, an update to Amazon Kinesis knocked AWS' US-EAST-1 region offline, disrupting businesses that relied on the service. Companies can invest in back-ups across regions to prevent disruption and operate amid an outage, but it's costly.
"Business IT leaders should provide a depth and breadth of service offerings while continuing to drive cost efficiencies," Alletto said.
Region concerns may not impact businesses experiencing low stakes in the wake of an outage, but can be concerning for organizations that need to be constantly online, such as healthcare facilities.
Factor regions into cloud purchasing decisions
When choosing a cloud region, decision-makers also consider what would best serve customers based on locations and any mandates the business will have to comply with as a part of infrastructure's locale.
"For those CIOs that are adopting multicloud and hybrid-cloud architectures, understanding how each cloud provider fits and meshes with each other is important," Moore said.
First, businesses should weigh if a multi-region deployment is right for them, according to Moore. A multi-region setup increases complexity and cost, and businesses can benefit from a clear understanding of how data flows.
"Observe and learn from some of the leaders in this area of multi-region footprint and leverage what they have learned and any tools that they have published to help them operate in a multi-region setting," Moore said.
For businesses interested in a multi-region setup, the idea of lock-in may hold them back. But most cloud customers understand lock-in is "occurring in a way that is no different than being locked-in to a relational database management systems," Raj Bala, research VP at Gartner, told CIO Dive in an email.
Still, IT leaders can evaluate for lock-in risk on a workload-by-workload basis, with consideration for the intertwined nature of some cloud workloads.
"Workloads are often deployed as a constellation of applications that are necessary for purposes of integration and concentration risk can occur within a given workload," Bala said. "Embrace the lock-in by deeply integrating with a cloud provider for some applications."
Going all in for some workloads helps businesses extract the most value out of a cloud provider, overshadowing the notion of lock in, according to Bala.