Hybrid Cloud Infrastructure

In the private cloud model, cloud infrastructure and resources are deployed on-premises and owned and managed by the organization. Private cloud requires a large upfront capital expense for equipment and software, a lengthy deployment, and in-house IT expertise to manage and maintain the infrastructure. It’s also expensive and time-consuming to scale capacity (because you have to purchase, provision, and deploy new hardware) and add capabilities (because you have to purchase and install new software). But private cloud provides maximum control over the computing environment and data, which is especially important—or even mandatory—if your company deals with highly sensitive data or is subject to strict industry or governmental regulations.

In the public cloud model, a company consumes compute, network, storage, and application resources as services that are delivered by a cloud provider over the Internet.
The cloud provider owns, manages, provisions, and maintains the infrastructure and essentially rents it out to customers, either for a periodic subscription charge or fees based on usage.
Public cloud offers significant cost savings because the provider bears all the capital, operations, and maintenance expenses. It makes scalability as easy as requesting more capacity, and it lets your company’s IT staff focus more on revenue-driving activities and innovation and less on “keeping the lights on.”
In public cloud's multi-tenant environments, your workloads are subject to the performance, compliance, and security of the cloud provider’s infrastructure. With Virtual Private Cloud (VPC) capabilities, you gain full control over your public cloud environment, including security and controls. VPCs give you the scalability of a public cloud and the security of a private cloud.

The hybrid cloud model represents the best of both worlds. You can run sensitive, highly regulated, and mission-critical applications and workloads or workloads with reasonably constant performance and capacity requirements on private cloud infrastructure. You can run less-sensitive, more-dynamic, or even temporary workloads (such as development and test environments for a new application) on the public cloud. With the proper integration and orchestration between the two, you can leverage BOTH (when needed) for the same workload. For example, you can leverage additional public cloud capacity to accommodate a spike in demand for a private cloud application (this is known as “cloud bursting”).​