Early cloud computing talk about the inevitability of a public compute utility has mostly given
way to more nuanced discussions. Cloud characteristics such as flexibility, rapid scalability, efficiency,
faster application development, and agility are as desired as ever, and are seen as essential
enablers for next-generation IT infrastructures. However, these infrastructures won’t come in
one shape and one size—and they won’t pop into existence overnight.
There are a number of reasons for this variety. For example, it’s well established that many
organizations plan to continue running at least some of their applications and workloads1 within
facilities that they control, rather than in a public cloud (for real or perceived reasons of security,
visibility, compliance, or control). It’s also true that many existing IT investments can’t and
shouldn’t transition to something new overnight; doing so would be neither practical nor costeffective.
As a result, most in the industry have accepted that clouds will often be hybrid blends
of public and private resources and heterogeneous infrastructures.
Hybrid clouds—specifically, open hybrid clouds that provide for application and data portability—
are also the destination of choice because the workloads that organizations run are likewise
hybrid. Hybrid workloads reflect the distinction between traditional enterprise workloads and
new-style workloads. New application architectures and technologies are better suited for highly
distributed, fluid IT infrastructures than was once the norm. But they also reflect the different
roles that workloads play within an organization.