The Lightcrest Blog

Let's talk about fluid computing, hyperconverged infrastructure, and hybrid cloud technology.

The Best Cost Model for Your Cloud: Rent Your Peaks, Own Your Valleys

Last update on Nov. 10, 2016.

One of the major benefits of the public cloud is the elasticity of the environment. You can spin infrastructure up and down on-demand, and you can shut everything down with the push of a button. This is great when your business is starting out, and you are still getting a feel for baseline traffic levels and load averages. But as you scale, the cost premiums associated with these services grow in tandem and your bill can become excessively large. 

The crux of the aforementioned problem is “convenience premium.” Public cloud providers charge a premium in exchange for immediate access to IOPs, storage capacity, virtual CPU, and other services. This is fundamentally how they make a profit - and it’s the key driver behind your costs as you scale your business up in their environment. While pundits still debate whether or not the cloud “honeymoon” is over, one thing is clear: hybrid cloud is the way of the future, both from an economics and performance standpoint. 

In a hybrid cloud, you can “own your valleys” and “rent your peaks.” This is the most cost-effective model because you stop paying rent on the majority of your infrastructure. By capitalizing a private cloud, you “own your valleys” - and drive all that cost out of your cloud bill. By leveraging a hybrid cloud, you “rent your peaks” - by bursting into public cloud services on an as-needed basis for temporary resources.

Owning Your Valleys, Renting Your Peaks 

Let’s start with an analogy. 

Imagine you have a new business that manufactures bicycles, and you need a factory. When you first open your factory, you don’t know how much demand you’re going to have - so you rent a tiny factory. Over the years, your business grows, and business starts to boom! But to accommodate the growth, you have to increase the size of your factory, and as a result your costs increase and your margins thin. 

At this point, you can continue to rent your factory forever. But if your business is continuing to grow, wouldn’t it be ideal to eliminate the rent associated with your factory so you can keep more of your hard-earned profits as you sell my more bicycles? Ideally, you could own your factory and stop paying rent on it. 

The beauty of the private cloud is, as the factory analogy, you can own it. Only it wouldn’t take 15-30 years to pay off like a factory; it would take two or three. 

This is a simplistic analogy, but it gets to the point. When you’re in a public cloud, you’re renting forever, even if your business is on a huge growth trajectory. This is why many companies leave public clouds at a certain scale. Infrastructure is so incredibly fast now that it is highly unlikely a hardware investment will be obsolete in 6 years as it pertains to your application. 

To own your valleys, you buy a private cloud. It can be financed and deployed in various ways, and usually, you can pay it off in 1-3 years. 

To rent your peaks, you consume temporary resources on a utility grid, or on public clouds. There is nothing stopping you from bursting into AWS or Azure while leveraging the economics of private cloud for your steady-state workloads. 

The combined implementation of “owning your valleys” and “renting your peaks” is called a hybrid cloud. 

Decreasing Cost and Increasing Performance with Hybrid Cloud 

With a hybrid cloud, you consolidate all your steady-state workloads onto infrastructure you own (or lease for a low cost) and burst into the public cloud for unique services like object storage and video processing.  In essence, it’s a “best of both worlds” approach to cloud computing, as you stop paying rent for the majority of your infrastructure, but you can still tap into the elastic services provided by the leading public clouds. 

For example, let’s say you’ve built a software solution for electronic medical records whereby your traffic is more-or-less growing linearly, and you’re steadily increasing your vCPU and block storage allocations over time. The most cost-effective way to handle this situation would be to forecast your growth over the next two years, architect a private cloud with room for at least twice that much growth, and then purchase it. By year two, your actual cloud costs would drop out of your bill, and you would be virtually rent-free. Meanwhile, any unexpected surges in traffic could be met through cloud bursting. 

By the time you hit year two in your growth trajectory, you can either add to your private footprint for future growth, or you can continue to remain in hybrid mode - depending on what’s the most cost-effective relative to your trajectory. 

Lightcrest and Kahu Compute Fabric 

Lightcrest empowers its customers to navigate the hybrid cloud landscape and minimize their cloud overhead. With the Kahu Compute Fabric powering their environments, customers get best-of-breed cloud technology that enables a turn-key private cloud without any of the datacenter overhead or required staffing costs associated with a home-grown solution.

 Since Lightcrest Datacenters are hybrid-enabled, customers immediately reap the benefits of leveraging the best of both worlds, whereby Kahu provides high-performance, single-tenant cloud environments combined with low-latency connectivity into leading public clouds. This powerful approach allows customers to shred their cloud costs while increasing their total cloud service catalog. 

At more than 50% more cost-effective than a traditional public cloud deployment, Kahu was built to provide small business and enterprises alike with reduced total cost of ownership and on-demand software-defined datacenter capabilities. To find out more about what Kahu can do for your organization or to speak to someone directly, don't delay - contact Lightcrest today for more information.