There’s a lot of talk these days about “the cloud”: data in the cloud, business in the cloud, security in the cloud… But what does cloud computing mean?
At its core, there is no universal distinction among the terms “cloud technology” and “cloud computing”. No matter how you phrase it, “the cloud” is a metaphor for any server-based infrastructure—whether you’re hosting data, applications, or both. What this means is that somewhere there sits a physical server hosting information that you then are accessing from another device, most likely remotely. This is similar to how servers host webpages and content on the internet, which is why many people also use the phrase “cloud” as a metaphor for the internet as a whole, and it’s also why it’s not inaccurate to say that cloud computing means hosting data or applications on the internet.
More and more businesses are beginning to utilize cloud computing (hosting their data and/or applications in a remote data center) over on-premise hardware. In fact, over the next 5 years, we expect to see 44% annual growth in public cloud workloads and only 8.9% growth in “on-premise” computing workloads. And 2014 is the first year in history that the majority (51%) of all workloads will be in the cloud.
There are several different types of cloud computing, and we’ve put together a run-through of them all and the variation among them.
SERVICE MODELS
Infrastructure as a Service (IaaS)
This is most basic type of cloud computing. Third party IaaS providers host servers, virtual machines and other network services in a remote data center. Frequently they offer other virtual computing resources as well. This service is typically charged under a utility model of pricing, where users pay based on the amount of resources they consume.
Desktop as a Service (DaaS)
Technically a subset of IaaS, DaaS providers host users’ desktops remotely. Users then access through Desktop Virtualization software, such as Windows Remote Desktop Connection. Centrality offers DaaS under a per-user pricing model. Regardless of consumption, our clients pay only based on the number of users.
Platform as a Service (PaaS)
The PaaS model delivers a computing platform for development. Oftentimes, this includes operating system, programming language execution environment, database and web server. With PaaS offerings such as Google App Engine, developers can develop software without investing in the capital for physical infrastructure, and often have automatic resource re-scaling so the user doesn’t have to upgrade hardware himself.
Software as a Service (SaaS)
With SaaS, providers offer their solution or software through a hosted platform they manage that runs the applications. This means users don’t need to install the actual software on their physical devices as access is generally through the web. SaaS generally has a scalable subscription-pricing model based on number of users.
DEPLOYMENT MODELS
Public Cloud
Cloud-based solutions accessed via the Internet and delivered through a shared environment. Often, they are with individual user accounts, such as gmail, Netflix, Dropbox and iCloud. Services may be free or offered as a pay-per-use model. Public Cloud solutions present greater security concerns than Private Cloud because of the greater scope of access.
Private Cloud
An in-house or outsourced cloud-based solution hosted in a closed data-center environment not accessible to the general public. Usually within a company’s domain and firewalls, private clouds are not readily accessible through the web. Thus, they inherently present fewer security risks than Public Cloud solutions.
