Skip to main content

The devil is in the detail

Multicloud
(Image credit: Everything Possible / Shutterstock)

There is now an extensive language around cloud deployments and within that lexicon, multi-cloud and hybrid cloud are two terms that are frequently used interchangeably.  However, these two terms don’t necessarily describe the same deployment strategy. 

Both hybrid and multi-cloud provide similar advantages and disadvantages, but with a few key differences that could be critical to any given organization. 

The devil is very much in the detail so it makes good sense to not only compare multi-cloud vs hybrid cloud, but also orientate that discussion so that a business can determine which cloud deployment fits its data requirements and goals in both the immediate short term and further into the future.

What is hybrid cloud? 

A hybrid cloud deployment combines a public cloud with a private cloud and/or on-premises infrastructure. The computing, storage, networking, and service resources are spread across multiple platforms, but with clear orchestration among them. The goal of a hybrid cloud deployment is to create a single, unified environment of applications and workloads that can easily move around as needed for performance optimization, failover, or to keep sensitive data close at hand.

To fully understand what makes a hybrid cloud different from a multi-cloud, businesses need to understand the small, but significant differences between a public cloud and a private cloud:

A public cloud is hosted by a third-party provider, such as Amazon Web Service (AWS) or Microsoft Azure, who controls and manages all of the hardware resources.

A private cloud uses data center infrastructure that may not be on-premises, but the customer business then manages and controls all of the hardware itself.  Alternatively the business may pay for a managed services provider to take care of it).

The key difference is that public cloud resources are shared by different businesses, whereas private cloud resources are dedicated to an organization.

There are benefits and drawbacks to both private and public cloud models, so many organizations choose a hybrid deployment to take advantage of the best parts of each. 

For example, if a business needs more control over physical and network security for a cloud database or application, it may want to keep it in a private cloud. 

However, building out a private cloud infrastructure can be very expensive, so that business may want to keep less sensitive resources in a public cloud to take advantage of reduced hardware and maintenance costs. 

A hybrid cloud deployment allows any organization to have the best of both worlds, matching resource to demand, while providing a seamless, unified environment.

What is multi-cloud? 

By comparison, a multi-cloud deployment combines multiple, public cloud platforms. There may or may not be any orchestration or unified management of a multi-cloud deployment, but it’s definitely recommended. 

Often, a business will end up with a multi-cloud deployment by default.  Many organizations unintentionally end up with a multi-cloud environment after years of ad hoc migrations to whichever cloud platform happens to best support a particular workload.  

There are many benefits to using a multi-cloud model in addition to being able to choose the best cloud provider for an application or service:

Firstly, deploying resources across multiple providers and platforms allows a business to avoid vendor lock-in, giving that business more leverage to negotiate contract terms.

Secondly, businesses can potentially use a multi-cloud deployment for disaster recovery and failover in case of an outage at one provider, or to provide load balancing for better application performance.

Thirdly, an organization can implement multi-cloud management tools and platforms to provide orchestration and management across all relevant public clouds, giving the business the same unified environment as a hybrid cloud deployment.

Fourthly, multi-cloud deployments improve a security posture, because if there is a data breach at one provider, only your data on that platform is at risk.

This can be improved further by distributing sensitive, regulated data across multiple compliance-certified cloud providers to further improve data security.

Multi-cloud vs Hybrid Cloud: Making the right choice 

The challenge of comparing multi-cloud vs hybrid cloud is that there is significant overlap in their advantages and disadvantages. In fact, they’re not even mutually exclusive terms – a hybrid cloud model can include multiple public cloud providers, turning it into a hybrid multi-cloud deployment. 

However,  to dissect which deployment model to choose, businesses should consider the following:

Cost. A private cloud is almost always significantly more expensive than a public cloud, because the business is paying for dedicated and, often, underutilized resources. Thus, a hybrid cloud deployment is typically more expensive than a multi-cloud deployment.

Orchestration. Though multi-cloud orchestration is possible, it tends to be a little more challenging than hybrid cloud orchestration.  This is because not all public cloud platforms integrate well together. However, it is worth keeping in mind that any orchestration across multiple providers is going to be difficult to manage whether a business opts for a hybrid or multi-cloud deployment.

Scalability. Scaling up a private cloud requires purchasing, installing, and integrating more dedicated resources. This can cause delays and demand a lot of extra resources. 

By comparison, scaling up a public cloud can often be accomplished with a simple code change, making a multi-cloud deployment model better for overall scalability.

Choosing the right cloud deployment for your data requirements 

The first consideration in assessing these types of deployment is that overall, choosing either a multi-cloud or hybrid cloud deployment will provide a business with greater flexibility than keeping all of its infrastructure and services in one location. 

The next logical step in choosing between multi-cloud or hybrid cloud is an in-depth analysis of specific, idiosyncratic data requirements and business goals.  This must also include an in-depth evaluation of potential public cloud providers to determine how a business could best migrate its workloads.

From here, the nature of the ideal cloud deployment should start to crystallize quite quickly – especially around issues of integration and orchestration, as well as the scalability to match future plans.   Keeping an open mind is vital to match the right technology, or mix of technologies, to a situation. But again, the success of this will depend on an analysis that looks at the detail of the business.

Andrew Davis, Sr. Director of Research and Innovation, Copado

Andrew Davis is senior director, product marketing, Copado and author of Mastering Salesforce DevOps.