What the latest cloud explosion really means

The cloud surge last week was like nothing I’ve ever seen. It reminded me of one of those sudden blasts of fog through the Golden Gate.

It all started with Cisco announcing it would invest $1 billion in an OpenStack-based “intercloud” to provide cloud services to large b-to-b customers. A day later Google slashed its cloud prices — Compute Engine by 32 percent, App Engine by 30 percent, Cloud Storage by 68 percent, BigQuery by 85 percent — and introduced Sustained Use Discounts that lower costs even further. It took Amazon just 24 hours to fire back with its own price cuts of 36 to 65 percent for its Simple Storage Service and 30 to 40 percent for EC2, among other reductions.

[ Which freaking PaaS should you use? InfoWorld helps you decide. | Stay on top of the state of the cloud with the “Cloud Computing Deep Dive” special report. | Stay up on the cloud with InfoWorld’s Cloud Computing Report newsletter. ]

The grand finale came in Satya Nadella’s press conference Thursday, where he announced Microsoft’s Office for the iPad and conjured a vision of a “cloud for everyone and every device.” Nadella made clear Microsoft wants to dominate the SaaS segment of the cloud, starting with clients distributed and administered from its Office 365 hub and ultimately embracing a new wave of devices not imagined yet.

What should we make of this fresh cloud explosion? Let’s take the three cloud flavors in turn.

1. The rush to IaaS

No one is sure exactly what Cisco has in mind for its OpenStack intercloud, but the idea of offering specialized cloud infrastructure services to, say, the telecommunications industry among other verticals is intriguing. If SDN means the days of high-end, high-margin networking equipment are numbered, Cisco had better lay the groundwork for high-end, high-margin cloud services.

As for the dramatic Google and Amazon IaaS price reductions, they still won’t convince enterprises to move big chunks of their existing infrastructure to the cloud wholesale. Large enterprises can’t tolerate that sort of dependency, and migration to Google or Amazon doesn’t absolve the customer from having to manage infrastructure, virtual though it may be. On the other hand, lower cost can only accelerate the usual enterprise IaaS activities: dev and test, disaster recovery, and more recently using IaaS as a platform for buidling customer-facing systems of engagement that need to change and scale quickly.

It’s a different proposition for businesses just starting out. Who wants to make risky capital investments in infrastructure? Cheap cloud capacity makes the all-in cloud practical, and the enticement of Sustained Use Discounts makes it less likely new businesses will eventually migrate their cloud deployments back to local infrastructure. Expect Google’s competitors to clone that pricing model soon.

2. App dev in the cloud

The most interesting part of the Google announcement actually involved App Engine, Google’s PaaS. Today, PaaS offerings like App Engine are used mostly by independent software developers to build, test, and deploy Web or mobile apps (Snapchat was touted at the Google announcement as an App Engine poster child). The problem with a public cloud PaaS, though, is that you’re limited to the languages, databases, and other services offered by the PaaS, which is why many devs who opt for the cloud build on an IaaS platform where you can build your stack can from the VM up as you like it.

Google’s new Virtual Managed Machines (VMMs) introduced a hybrid approach. You build, deploy, and manage the core of your app or service on App Engine, where you can now also manage VMMs containing other components of the app or service deployed on Compute Engine. In addition, Google made many App Engine developers happy by announcing integration with both Git and GitHub.

PaaS is the hottest area of the cloud right now. Enterprises are actively exploring public cloud PaaS for building new, customer-facing applications — many are looking at on-premises PaaS as well. Inside the enterprise data center, PaaS is poised to replace the application server, ushering in scale-out architecture and support for multiple languages.

3. SaaS redefined

Technically, SaaS is defined as multitenanted software maintained by the provider with a browser-based client, like Salesforce or Google Apps. But the user experience needs to come first. Using Office for the iPad is vastly better than trying to use Office Web Apps on the iPad, for example.

If Microsoft has really changed its tune and plans to support all platforms equally instead of saving the best user experience for Windows, then Office 365 and Azure could together become a SaaS powerhouse. Not necessarily SaaS in the Web app sense, but as a cloud distribution center for software by subscription, with Exchange, OneDrive file storage, Azure Active Directory for user identity, and more all maintained in the Microsoft cloud.

Ultimately, I can also see Microsoft expanding its line of enterprise apps, modeled along the lines of Microsoft Dynamics CRM, which is a classic SaaS app with a browser-based client. That’s the new software model: native clients where they’re needed for the user experience, JavaScript where it isn’t, and back-end functionality in the cloud.

To be clear, the events of last week do not indicate that we’re getting any nearer to putting “everything in the cloud.” But it’s getting clearer what should and shouldn’t be there. And what should be there is getting closer to its destination.

This article, “What the latest cloud explosion really means,” originally appeared at InfoWorld.com. Read more of Eric Knorr’s Modernizing IT blog. And for the latest business technology news, follow InfoWorld on Twitter.