Editor’s Note: The final section of our cloud computing tutorial defines what a hybrid cloud deployment is and helps network managers understand the associated pros and cons. We examine how the network is affected by a hybrid cloud deployment and how the network needs to evolve to ensure optimum performance.
Don't miss any of the articles in this tutorial on network planning for cloud computing:
A hybrid cloud is a mix of both private and public clouds. This allows the organization to gain the benefits of cloud computing only where it makes sense. A common scenario is for an organization to keep its data in its own data center and then use a public cloud service to perform whatever computation tasks are required. A hybrid cloud allows an organization to leverage its current investments in compute infrastructure and augment it with a cloud-based service. Because this model allows organizations to migrate to cloud computing at their own pace, hybrid clouds will be the path most will choose to begin their cloud deployment.
Because the hybrid cloud model allows organizations to migrate to cloud computing at their own pace, hybrid clouds will be the path most will choose to begin their cloud deployment.
At first glance, it appears that hybrid clouds provide the perfect mix of private and public, but that’s not necessarily the case. Ideally, in a hybrid environment, a portable computing resource, or workload, would be able to move seamlessly between the organization’s private and public cloud service. But the network will need to play a key role to secure and optimize this movement as it traverses the Internet. To do this, the network needs to provide the necessary security as well as QoS.
Securing the workload is difficult to do in a hybrid environment because the organization will lose full control of the workload when it traverses the Internet. Security needs to be addressed at all layers of the cloud environment, including the network. Most traditional compute resources are secured by external systems. In a hybrid cloud model, the workload itself will need to maintain a state of security while being migrated from the private to public cloud. The network needs to know when the workload is moving from the private to the public cloud and then apply the security to the workload as it leaves the private cloud. The best practices put in place in the public cloud model should be applied to the public portion of the hybrid cloud.
In a hybrid environment, QoS needs to be applied at the workload level to prioritize it over other network traffic. Without prioritization of the workloads, it’s nearly impossible to understand how long the actual workload migration will take, and a service-level agreement (SLA) cannot be determined. To accomplish this, the network and virtualization layers need to be unified to ensure that when the workload moves, all the necessary network configuration parameters move with it. This is the basic concept behind the vision of unified computing, which is a relatively new concept that will continue to evolve as organizations build out hybrid clouds.
Regardless of whether the organization chooses to deploy a private, public or hybrid cloud, it’s critical that the network play a key role in securing and delivering the cloud resource to ensure optimum performance and maximum return on investment.
About the author: Zeus Kerravala manages Yankee Group's infrastructure research and consulting. His areas of expertise involve working with customers to solve their business problems through the deployment of infrastructure technology solutions, including switching, routing, network management, voice solutions and VPNs.
This was first published in December 2010