The Cloud. The term in tech is as amorphous as a real cloud. Clouds come in various types: public, private, hybrid, there is even a fog. But what exactly are they all and how do they benefit organizations? We’ll do a quick dive into what these are and discuss which one we feel has the most benefit.
Public cloud is the term that is most familiar as it is the one we use everyday. Most likely you are using the public cloud many times a day with Google’s email, maps, search, etc. When you are streaming Netflix you are consuming multiple services from Amazon under the hood. The public cloud is just that, the infrastructure and services that are publicly accessible from the internet.
Private cloud represents servers and services that are proprietary to a single organization. This form of cloud is still elastic and highly scalable but is internally managed and operated by the organization that created it.
Hybrid cloud is the peanut butter in your chocolate, it can be an intermixing of public cloud services, but is more typically a blend of private cloud (or on premise) with public. From here we will discuss why organizations would want this mix when the drive to public cloud seems so ubiquitous. Why would you want to maintain a hybrid cloud model when the public cloud can take the burden of infrastructure and DevOps off your plate? The quick answers are cost and ownership and innovation.
Many services in the public cloud are inexpensive, even on scale. Leveraging serverless compute functions or an image recognition service that you pay for true on-demand use makes a ton of sense. The infrastructure is battle tested and it manages your spiky traffic. Image recognition isn’t the central focus of many IoT projects and it doesn’t make sense to train/hire a team to tackle that problem. However your application database is always running and its footprint is so large that the monthly costs in the public cloud end up being far greater than if you brought it in house.
If AWS goes down, much of the internet goes down. While this isn’t a common occurrence it does happen. And when it does, having only a public cloud presence leaves you dead in the water. The inverse is true if you are solely managing infrastructure internally. Having the internal and external redundancy gives your organization a massive leg up on resiliency and fail over. Managing mission critical systems on public cloud is also a black box. You trust your provider but sometimes performance goes sideways and only after much investigation it’s discovered to be a noisy neighbor. For the infrastructure that means the most to an organization it makes a lot of sense to bring it internally. You have full view into patches, hardware and performance.
There is a great sense of ownership and pride in building something yourself and maintaining it. Building everything also means you need to fix it, keep it alive, and retain the domain knowledge that made it. Back to our image recognition service, it may not make sense to roll your own. You need the feature but it’s not critical to your IP plus the development time and costs are out of reach. Having access to public services allows your team to quickly add new features and pull them into what is proprietary and important for your organization.
The final statement is that having a hybrid approach to your infrastructure allows you to manage costs, innovate quickly, while keeping what is most important managed internally.