Cloud

Why enterprise cloud projects fail: three topics not to PaaS on

We will once again tackle the topic of polyglot versus best of breed PaaS at VentureBeat’s CloudBeat conference in a few weeks, a topic that only touches a sliver of how PaaS should be talked about.

Unfortunately, PaaS, as defined and delivered by most, has generally been a fisherman’s tale; a story of over promise and under-deliver. If the industry faces down a few tough topics, PaaS can be back on track with a truly transformative vision.

Here are three topics the panel won’t have time to cover, but should:

The next five years will be shaped by private, not public PaaS

Surely, anyone reading this is saying “Are you kidding? Every developer I know uses public PaaS and it’s the future!” While this might be true for independent developers, this is not true for professional developers in the enterprise. The fact is, custom application workloads running inside enterprise datacenters are not easy targets for moving to public PaaS. They are mired in regulatory and security concerns, have capabilities hard-coded for their enterprise environments, and require physical proximity to massive datasets.

This puts enterprises in the tough spot of having to “pass on PaaS” for a huge number of their workloads. Lucky for them, Private PaaS gives enterprises the same agile operating model as public PaaS, but without the adoption hurdles. This will drive workloads to PaaS, and help all of us. PaaS environments are starting to focus on hybrid cloud outcomes, which means these workloads, can then “migrate” from private resources to public ones through the PaaS. The result is that private PaaS is the easy to adopt, high return option that will ultimately drive public PaaS adoption by acting as the logical layer that pushes workloads to public environments.

PaaS as we know it is dead: cloud architecture “enabling” platforms will rule

To date, PaaS has been a story focused on easing the deployment and management burdens of applications. Essentially, standing up “old” stacks in an abstract way thereby making easier for developers to get their apps deployed and scaled since these distracting workflows are commoditized.

While valuable, this is a tactical view of the cloud. Cloud has driven a fundamental change in how applications are written and architected – so much so that we now refer to “cloud applications” as a category.

What is a cloud application? Typically, it’s any application that is some combination of multi-tenant, mobile, distributed, or web-scale. These “architecture patterns” are ridiculously hard to build. What’s worse is that *most* PaaS’ (there are some, but not many) do absolutely nothing to help developers build cloud applications. Kind of ironic that cloud architected platforms don’t help developers build cloud apps, isn’t it?

CloudBeat 2012CloudBeat 2012 assembles the biggest names in the cloud’s evolving story to uncover real cases of revolutionary adoption. Unlike other cloud events, the customers themselves are front and center. Their discussions with vendors and other experts give you rare insights into what really works, who’s buying what, and where the industry is going. CloudBeat takes place Nov. 28-29 in Redwood City, Calif. Register today!

Some vendors with strategic vision, however, are reshaping the landscape with PaaS offerings that can enhance web and SOA applications and “inject cloud DNA” into those applications. The result? Developers can build next generation application architectures – like multi-tenant architectures that salesforce.com put on the map – with little to no effort or skillset change. These “cloud architecture enabling platforms” will define a new gold standard in PaaS, killing off any technologies whose “vision” is to deploy applications.

Near term, PaaS will end up being a cost cutting war

A corollary to my point above is that PaaS vendors have lost site of a bigger vision for cloud platforms and will soon find themselves competing on price.

Most PaaS technologies will converge on common, deployment oriented feature sets where incremental increases in “how fast something deployed” will be the biggest differentiating factor. As a result of the lack of unique differentiation, price cuts will ensue. This is a direct reflection of a near sighted view of the market, and if vendors don’t focus on moving PaaS from deployment to next generation application development, they’ll find themselves in an unfortunate business model.

The crew on the PaaS panel at CloudBeat has outlined a great discussion, but if we had 30 more minutes, I’d attack these larger topics in the market. As PaaS spending continues to grow, taking the aforementioned challenges head-on will guarantee that PaaS becomes the transformative layer it was intended to be.

Sinclair Schuller is the CEO and Co-Founder of Apprenda.

He has deployed more than a dozen, paying enterprise private PaaS solutions, just not “deployed” small apps. Sinclair therefore has a unique perspective on what enterprises are considering when look at next gen apps and modernizing existing app portfolios and how PaaS (public, private, hybrid) need to build, support AND deploy these complex app architectures.

Web services diagram: Shutterstock

blog comments powered by Disqus