Maximizing Cloud Optionality

April 2, 2015 Coté

featured-PCF-PWSA rare opportunity to get excited about enterprise licensing!

We’re making new licensing terms available to customers that allow them to choose between running their applications in our public Pivotal Web Services platform or their own Pivotal Cloud Foundry at no additional charge. Customers of Pivotal Cloud Foundry can choose to run their applications in a public, multi-tenant environment or their own single-tenant Pivotal Cloud Foundry based on timelines and risk profiles. We’ve also announced support for Pivotal Cloud Foundry in an Amazon Virtual Private Cloud on AWS. This maximizes flexibility for customers, giving them options that can speed up project time and also layering in public cloud elasticity as needed to their applications.

We’ve been running Pivotal Web Services for some time, and of course Pivotal Cloud Foundry has established a strong foothold in the private cloud world over the past year. However, because we’ve been billing them as separate offerings we and our customers have been treating them as different services. Now with some small licensing changes we made last week, customers of Pivotal Cloud Foundry can move between the two without incurring extra costs. Equally importantly, Pivotal Cloud Foundry customers now have all types of cloud at their disposal, helping them get the multi-cloud capabilities: they can run in full-on, public cloud multi-tenant mode, single-tenant “public cloud” in an AWS VPC, or “private cloud” on their own hardware. The chart below illustrates all the runtime options:



“There’s Nothing I Love More Than Sorting out Licensing Terms!” Said No One Ever

When the overall mission is to shrink the time it takes to get operating, you can imagine that the last thing we want to do is add add friction to that process from fine print and licensing concerns. So, our goal with these changes is to make the pricing and licensing issues go away for customers who want access to both modes of running Pivotal Cloud Foundry.

With these licensing changes, customers can choose to run their application instances in Pivotal Web Services or Pivotal Cloud Foundry with no difference change in pricing. And that includes everything: there’s no additional charge for the underlying cloud infrastructure (we run Pivotal Web Services on Amazon Web Services). You get the same 24/7 support wherever you’re running your applications. The only thing you “give up” when running in the Pivotal Web Services is some of the operations concerns and capabilities that you’d have if you were running your own platform; for many people who use Pivotal Web Services, this is of course a bonus. They don’t want to be managing cloud, rather, as Dr Nic said, “They just want to be building apps.”

Focus on Your Applications, Not the Cloud Taxonomy Dialectic

One of the more onerous discussions in recent years has been the never-ending, maddeningly peripatetic discussion around how to categorize clouds. Depending on how you slice up public, private, hybrid, further divided by who manages it and if it’s single or multi-tenant…you can get into some Vulcan chess levels of dimensionality sophistication that kind of go nowhere. (And also, whatever happened to community cloud? Add that one into the list for more taxonomic fun.)

Clearly, having access to a highly scalable, public cloud to handle surges in traffic is attractive and handy capability to have for those Pivotal Cloud Foundry customers who would like to use Pivotal Web Services. One of our customers, the Sundance Institute, used Pivotal Web Services for just this scenario recently. However, the strategic implications of this licensing change for our customers are more than just being able to offload seasonality-driven workload demands to Pivotal.

As I work more closely with Pivotal Cloud Foundry users, I’ve found that they’re less interested in nailing the correct taxonomization of their cloud project and more interested in two capabilities cloud can afford:

  1. Speed and agility—getting up and running quickly to start developing their application, even experimenting with their application by having real, actual people start using it. Few things are more of an application development buzzkill than “step one: build a private cloud” and this licensing sharing eliminates said buzzkilling. Line of business executives learn to love cloud because they can quickly develop and stand-up applications. In fact, our developers in Pivotal Labs often work with customers to do just this: using Pivotal Web Services as their initial rapid application development and delivery platform while they’re standing up their own clouds. They then move the workloads in house after getting their cloud platform up, or perhaps they keep it in Pivotal Web Services, depending on their needs.
  2. Maximizing optionality—once the application is “ready” enough to be deployed widely into production, our customers often want to maximize the deployment options they have available. Sometimes you’ll want to run in full on public cloud that’s multi-tenant, sometimes you want to run on-premises on your own hardware that you managed…and sometimes any number of gradients in between. A large part of attaining this optionality is the portability that comes from writing to a platform that can run on multiple types of infrastructure, like Pivotal Cloud Foundry, instead of being directly dependent on the cloud infrastructure. Of course, simplifying how you pay for that portability is the other part: with this new license entitlement option, there’s no need to worry about additional charges as the applications are moved around.

The first point is pretty self-evident: getting a Pivotal Web Services account setup and pushing an application only takes a matter of minutes, depending on how quickly you can type on the command line and push your first application. That deployment option isn’t always desired in production by companies who want, or need, more control over their platform than a multi-tenant, public cloud service provides, but many of them would like to start the process without blocking on standing up their own clouds first.

Indeed, on this second point, during my analyst work, I noticed that the ever popular term “hybrid” was often used to represent this goal of not really being nailed down too early on the exact type of cloud people wanted to use. In survey questions, customers would consistently say they wanted hybrid cloud which I soon took to mean not a formal, technical definition but instead more of a strategic goal of “I want as many options as possible.” You can see the step jump in interest in this chart from a 451 Research study on the topic last year:


It’s clear that users of cloud—our customers—want access to all sorts of runtime options, not just one “type” of cloud. And in the type of large organizations we tend to work with, they usually want all types! This new licensing scheme definitely gives that optionality: when it comes to licensing, don’t worry about where you’ll deploy your application instance right now if you don’t need to. Looking forward, several years, who knows if things will change the same or narrow down to one of the types of cloud (public, private, single-tenant, multi-tenant, a combination of all of the above, etc.). Regardless of that, what we at Pivotal are doing strategically here is removing that worry from our customer’s bucket of things to stress out about. We’re more interesting in making sure our customers can focus on the real value: writing and running the applications that help make their businesses better.

About the Author


This Month In Data Science: November 2015
This Month In Data Science: November 2015

With the Presidential race heating up, the increasing importance of data science within the candidates’ cam...

Cloud-Native On The Meetup Circuit
Cloud-Native On The Meetup Circuit

Starting Monday, November 30, the Cloud Foundry and Java Spring teams will be visiting Minneapolis, Columbu...