You may have already realized it. You’re going to have to adopt the cloud at some point. And as that point comes closer, it’s time to get a realistic idea about what the cloud offers so you’ll know how best to take advantage of it in your digital transformation.
In this post we’ll cover the basic cloud computing deployment models, strategies, and the various deployment architectures in use today to give you a primer on how the cloud is structured at the fundamental service level.
Read on below, or use the links to jump down to the individual sections on:
Before we start looking at the specifics of cloud adoption strategy, let’s take a look at the various shapes of cloud computing that are available for consumption today.
There are three major public cloud providers, also known as hyperscalers: Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform (GCP). The meteoric rise of these three providers over the last five years is a clear testament to the upward trend of cloud adoption. AWS remains the market leader in terms of market share and number of services offered. But Azure has established itself as a clear number two that’s closing the gap all the time. Azure is the popular choice of cloud for enterprises and large organizations. Google Cloud Platform is the fastest growing of the big three and is considered to be an innovation leader.
Each of these cloud providers offers near-identical services, but the nature of those services can be broken down into three different computing models: infrastructure as a service (IaaS), platform as a service (PaaS), and software as a service (SaaS). Let’s take a look at each of these models to see what they offer and how you can leverage them.
In general, there are three main cloud computing models available from the public cloud providers:
Considered as the most basic building block of cloud IT, IaaS provides cloud-based, on-demand, self-serviceable access to compute resources: either as virtual (consumed as virtual machines) or dedicated hardware, storage, and networking resources. This is often the first foray into cloud computing for many customers.
While these resources look and feel like some familiar on-premises technologies, unlike on-premises, the underlying infrastructure is managed by the cloud platform provider rather than the end user, who is only consuming the resources (such as the virtual machines) on a pay-per-use basis.
Popular examples of IaaS compute offerings include Amazon EC2, Azure Virtual Machines, and Google Cloud Compute Engine.
PaaS takes the concept of IaaS to another level. PaaS extends the abstraction of the underlying infrastructure all the way up to the operating system level for the end users.
This means the end users don’t have to manage the underlying infrastructure virtual machines, or the operating systems running on them. Instead, users focus their time and energy deploying and working with their applications, which are built on top of the PaaS offering.
While this somewhat reduces the flexibility at the infrastructure level, it helps end users quickly consume specific application services increasing the speed to market, while reducing operational complexity.
Examples of PaaS offerings include Amazon Elastic Beanstalk, Google App Engine, and Azure Web Apps.
The SaaS model further elevates the abstraction level all the way up to the software itself, where the end user is directly given access to the software to be consumed as a service. The user is not involved in procuring or managing any underlying software or infrastructure components.
Typically, all SaaS applications are cloud based and are accessible via an internet browser or an app. The end user only has to worry about the data held on the application as everything else underneath—from the application itself to the underlying operating system and virtual/physical infrastructure—is managed by the cloud provider.
Popular SaaS offerings include Microsoft 365, Salesforce, and Google Workspace (aka, G Suite).
There are clear benefits to each IaaS, SaaS, and PaaS that are available in the cloud for fully automated, self-serviceable consumption. Choosing which—or which mix of the three—is right for your organization is a key part to your cloud adoption strategy.
In addition to the three cloud computing models covered above, it is also important to understand the main cloud computing architectures available.
There are four main types of cloud computing deployment models in wide adoption today.
Often privately owned and managed by single entity (i.e., an enterprise organization) for their internal consumption, a private or on-premises cloud is an evolution of the on-premises IT model with some characteristics of the public cloud added, such as self-serviceability and semi-automation for resource provisioning and consuming.
It is important to note that a private cloud can be internally hosted within a customer's own data center, or externally hosted such as in a co-location facility, or even hosted at a local service provider as a managed service offering. Many enterprise customers tend to continue using such private clouds running in their own data centers, that are managed by their dedicated IT teams who handle procurement, provisioning, and the day-to-day administration of the infrastructure platform.
While this model of deployment provides benefits such as granular control, security, and flexibility, it usually comes at a high CAPEX (capital expenditures) cost.
The public cloud is owned and operated by large cloud service providers (aka hyperscalers, covered above) as a geographically distributed, global, often shared IT infrastructure, with logical separations for secure multi-tenancy. Typically accessed via the internet, cloud service providers design and build these platforms for scale and ensure logical security boundaries for customers to consume these environments securely, in a self-serviceable, fully automated manner.
Most importantly, the public cloud provides its users with the ability to pay for the services on a per-use basis, marking a significant shift from the traditional IT model from a CAPEX model to an OPEX (operational expenditure) model.
The public cloud provides various benefits such as easy scalability, high reliability, availability, and cost efficiency while also providing end users with access to multiple computing models (as described above), which aren’t all usually available on other deployment models.
A hybrid cloud deployment architecture consists of a mix of both public and private/on-premises cloud deployments. By combining the two, customers can benefit from the capabilities of both models and provide a more tailored IT solution for each line of business (LOB) in their organization, based on their specific business needs.
Hybrid cloud management is highly flexible and often enables customers to meet stringent requirements such as security and data sovereignty, without losing the ability to benefit from the innovation, cost efficiency, and scalability of the public cloud platforms. However, hybrid cloud strategy can come with a level of operational complexity that makes it only suitable for specific customers such as enterprise organizations.
Since hybrid architectures have access to both traditional and public cloud resources, they have a lot more flexibility to meet the traditional and new business requirements. Many enterprise customers who are starting to consume public cloud IT solutions as they continue to use their existing on-prem IT environments are driving the popularity of the hybrid cloud architecture model.
While it can be argued that the multicloud architecture is a specialized form of hybrid cloud deployment, the multicloud model is predominantly focused around combining multiple public cloud platforms (two or more public cloud platforms at minimum), with or without private clouds. This approach provides an organization the ability to distribute applications or IT services across multiple cloud platforms managed by multiple cloud service providers with a heterogeneous architecture and governance model across the board.
While multicloud deployment enables customers to consume best-of-breed technologies from each cloud platform, the main aim and benefit of the multicloud deployment model is to eliminate the reliance on any single cloud provider. However, multicloud models can come at the cost of added operational complexity due to the need to design applications that work across multiple cloud platforms and with different proprietary technologies.
Customers can minimize such complexity by having a common application layer (such as Kubernetes) and/or a common data storage and management layer (such as NetApp Data Fabric) that spans across these multiple cloud platforms.
Now that we’ve covered the basic cloud computing deployment models, it’s time to consider how you’re going to put it into use. What’s your cloud adoption strategy going to be? To find out, read more about the different cloud adoption strategies here, where you’ll take a deep dive into each of the available cloud adoption models and learn how to choose which one will best suit your needs.
Cloud adoption presents significant challenges that need to be overcome, such as managing large-scale migrations, changing systems and processes, controlling costs, and gaining visibility across deployments. These are challenges that NetApp Cloud Volumes ONTAP and Cloud Manager can help you solve.
NetApp Cloud Manager acts as your data estate platform and management layer on top of on-premises and public cloud storage resources, helping you to build, protect, and govern your data. Using Cloud Manager, users have a single pane of glass that can extend to all deployments, across any of the big three clouds.