Hybrid cloud architectures sound complex. In this two-part blog series, we seek to eliminate some of the confusion by digging deeper into the common characteristics of hybrid cloud architectures. Over the course of this series, we’ll define the hybrid cloud and highlight its five key architectural attributes to give you and your business a better understanding of why a hybrid cloud is the best fit for your workloads.
In this first installment, we talk about what the hybrid cloud is and why you should consider it, and highlight the first two common characteristic of hybrid clouds: legacy application architectures and large-scale cloud consumption.
There has been a ton of buzz in the industry lately around hybrid cloud. From enterprises developing multi-vendor cloud strategies to hosting providers offering products that take advantage of different infrastructure types, there has been a definite shift towards hybrid clouds as the preferred IT hosting model.
But what is hybrid cloud? Is it using a public cloud with a private cloud? Or is it using dedicated hosting with a public cloud? Or could it be using an on-premise private cloud combined with an off-premise dedicated hosting provider? The answer is yes! Hybrid cloud could be any combination of using a public cloud, private cloud or dedicated hosting. With a hybrid cloud, you can spread your workloads across each of these different environments to run them where they’re the best fit for your application.
1. Legacy Application Architectures
One of the most common attributes of a hybrid cloud architecture is that it enables a business to continue to leverage legacy application architecture principles and preserve their previous investments. A large amount of capital may have already been invested in the form of development work, software procurement or hardware. Leveraging bare metal or private VMware is often a great way to continue to use these legacy application designs that enable you to solve for high availability with infrastructure layer solutions such as shared storage clustering, VMware HA and DRS.
For example, let’s look at something like an Oracle eBusiness Suite, which is tightly integrated with many organizational billing systems. These systems are often designed around high available bare metal server architectures with hardware specific licensing models deeply ingrained. With the Rackspace hosted portfolio you could easily deploy an Oracle RAC environment on dedicated servers and leverage the Rackspace RackConnect hybrid automation solution to have intra-datacenter low latency connectivity to on-demand cloud resources. To get the best of both worlds, you can abstract new greenfield applications or web and application layer compute workloads into the public cloud while still maintaining the high performance connectivity to your backend or legacy systems.
2. Large Scale Cloud Consumption
A second commonly overlooked use case for hybrid cloud architecture is the efficiency gains for large scale cloud consumption. This model makes sense even for “pure” cloud application deployments, and makes great use of an open source platform like a private OpenStack cloud paired with Rackspace’s RackConnect, which delivers low latency connectivity to Rackspace’s OpenStack powered public cloud resources. One of the major benefits of cloud computing that most everyone is familiar with is its on-demand utility-based nature that provides rapid and nimble elasticity. What is not always considered is the premium that consumers must pay per compute unit for this fantastic benefit. If you’re running large scale applications in a public cloud without any dedicated resources, you’re most likely spending more money than necessary to achieve your goals. Depending on your application, there is a point in scale (it could range from a few large cloud servers to many small cloud servers) where the cost of running the same amount of compute resources on a dedicated private cloud becomes more cost effective than in the public cloud. The private cloud also delivers the additional benefits around performance and security. We advise customers with large cloud footprints to migrate their always-on persistent cloud footprints to an appropriately-sized dedicated private OpenStack cloud footprint and leverage public cloud for their peak resource demands. As growth trends continue, you can add more compute nodes to the private cloud in a controlled manner without sacrificing the on-demand bursting capabilities to the public cloud. In a Rackspace-hosted private cloud you do not pay per instance running on your private cloud, you only pay for the hypervisors you lease and the level of support you would like to layer on top.
Blog: How HubSpot uses the Hybrid Cloud
Resource: Rackspace Private Cloud
That’s it for the first installment, be sure to check back next week as we discuss the next three common characteristics of hybrid cloud architectures: performance optimization; security requirements; and corporate policy, compliance and SLAs.