So many clouds, so many cloud choices. Back in the early 20th century, General Motors bragged, "a car for every purse and purpose." Today, there seems to be a cloud for every possible use case.
Cloud proliferation is a growing concern. The cloud's advantage is simplicity and agility, and multi-cloud has become the preferred state to optimize the cloud experience, observes Bryan Thompson, vice president of HPE GreenLake cloud services. "However, with too many clouds, enterprises trade simplicity for management complexity and higher costs."
A Danger in the Clouds
A specific cloud can offer unique functions, but becoming overly reliant on those capabilities may actually introduce complexity and inadvertently create a more costly cloud infrastructure, Thompson notes. "Multiple clouds can create an inconsistent management, integration, and performance experience that necessitates unique tooling and processes for each cloud," he warns.
Runaway cloud deployments can also lead to reduced productivity and growth. "Most businesses run what they consider killer apps, which pay the bills," says Leo Leung, vice president, Oracle Cloud Infrastructure and Oracle Technology. "Those applications should operate on a cloud optimized for that sort of work."
In the pre-cloud era, IT leaders might pick one software stack for databases, another for desktop applications, and yet another for graphics production. Leung suggests that today's enterprises should look at which cloud is best for each critical workload. "Picking one cloud for everything is riskier than adopting the right cloud for the right job," he notes.
The Numbers Game
The number of clouds an enterprise uses isn't as important as having clouds that meet current workload needs. "If you have 100 applications that follow two patterns of characteristics, then you likely need one or two clouds," Thompson says. "But if you have thousands of applications with distinct requirements, or need a proprietary cloud service, then you should consider more clouds."
