Most recently in this series, we took a look at the different available cloud service models of Software as a Service (SaaS), Platform as a Service (PaaS) and Infrastructure as a Service (IaaS). At a high level, the cloud service models describe what is being delivered to consumers. In this post, we’ll explore cloud deployment models. In my opinion, in contrast to cloud service models, cloud deployment models describe more of how the different cloud services can be deployed. Before we get into the individual cloud deployment models, I think it is import to revisit the characteristics of cloud computing, as defined by NIST. Remember, the cloud isn’t just “somebody else’s data center”. For brevity, I’ll just list them here, but please visit the previous post on cloud characteristics.
- On-demand self service
- Broad network access
- Resource pooling
- Rapid elasticity
- Measured service
Now that we have had our cloud characteristics refresher, let’s jump into the different available cloud deployment models.
As I learned from Cloud Bart, private cloud is all about exclusivity. A misconception I had before I started this learning journey was that private cloud meant it had to be a data center that you, as the consumer, owned and operated. While that can be the case, it does not have to be that way. As alluded to in the first sentence of this section, private cloud is less about location and more about the exclusivity of services. At a high level, private cloud essentially means that you are not sharing resources and storage with other customers. You also maintain control of your data. This also means that private cloud will likely be the highest cost deployment model solution for you as the consumer. The private cloud model makes sense for organizations that need to abide by certain regulations that require that level of exclusivity.
Public Cloud is very similar to private cloud, just without that exclusivity factor. With the public cloud deployment model, consumers receive hosted services from a cloud service provider. The resources the consumer leverages in this model are pooled by the CSP and distributed as necessary among the different, unaffiliated consumers of the CSP’s offerings. Consumers leveraging public cloud also have less control over where their data is stored. In the cloud assessment phase, consumers need to understand their requirements to make sure that that their data can be stored in the public cloud model. Public cloud can typically be thought of as the least expensive cloud deployment model.
It is probably not the best explanation, but I think of the community cloud deployment model as somewhat of a hybrid between private and public cloud. The community cloud model is multiple organizations with similar requirements or regulations going in together on a cloud solution. The cloud deployment and resources are exclusive to the organizations in the community cloud. In my opinion the biggest benefit to community cloud is that the participating organizations share the cost of the cloud deployment.
Hybrid cloud seems to be a common and popular scenario for different organizations. The concept of hybrid cloud is that a consumer leverages services in at least two of the different cloud deployment models. For instance, an organization could run a government regulated application in private cloud, and a public facing website in the public cloud. Hybrid cloud is sometimes referred to as the “best of both worlds” model.
A key takeaway that I have gotten from my Cloud Essentials+ studies is the contrast between cloud service models and cloud deployment models. From my interpretation, cloud service models deal with what is being delivered from a cloud service provider, and cloud deployment models focus on how those services are being delivered or deployed. The defined cloud deployment models are private, public, community, and hybrid.
2 thoughts on “Cloud Essentials+ Journey – Cloud Deployment Models”
I also had the same misconception about private cloud until you pointed this out – thank you for sharing what you’re learning.
Thanks for reading, and the feedback!