This is my first blog post as Editor-in-Chief of the upcoming Modern Infrastructure e-zine, filed to you directly from VMworld 2012 in San Francisco.
On my way to breakfast this morning, I grabbed a complimentary paper copy of the New York Times (I know, how quaint…). Page one featured a glowing story about Amazon Web Services, whose goal is to give “anyone on the planet an almost unimaginable amount of computing power,” the paper giddily reported.
Judging from the show floor, VMworld’s 23,000 attendees and 250-plus vendors haven’t read the article, or at least, don’t seem to think it applies to them. There, every IT vendor on the planet except Amazon (which is not surprisingly, not exhibiting) is busy regaling eager attendees with demos of their latest “engineered systems,” private cloud management stacks, flash arrays and backup software. These are all products predicated on robust in-house data centers teeming with dedicated infrastructure – but is that really the way things are headed?
VMware has built out an amazing technology and ecosystem, but let’s face it, the conversation has moved on. It’s AWS, not VMware, that has captured the hearts and minds of developers and startup owners everywhere, and its dirt-cheap costs, global presence and plentiful add-on services are increasingly driving enterprise business owners to the platform.
That has tremendous implications for VMware and its users. While there’s an enormous amount of on-premise, non-cloud infrastructure out there that needs to be configured and maintained, that estate will dwindle over time as more and more workloads are born on or migrated to the cloud – and from this vantage point, it doesn’t look like that cloud will be based on VMware, said Sacha Labourey, CEO and founder at CloudBees, a Java PaaS startup.
Amazon is not only the leading cloud player today, the race isn’t even close. There is some desire for Amazon alternatives, but the depth of its services and the speed of innovation “are like a magnet” that draw more and more customers in its fold, said Labourey.
VMware’s cloud efforts like vCloud and Cloud Foundry, meanwhile, are perceived to be expensive and moribund, and the company is stymied by its existing revenues and partnerships from going all out on cloud.
VMware is stuck between remaining in the past as a software vendor, or moving forward and becoming a service provider, Labourey said. The former has a terrific revenue stream and lots of inertia, “but what’s next?” and the latter creates conflict with its partner ecosystem that is hard to swallow. “It’s a tough move to make.”
For now, VMware insists that it will stay the course, despite rumors that it will jump in to the IaaS market with Project Zephyr. Asked about Zephyr at a press conference yesterday, Bogomil Balkansky, VMware senior vice president for cloud infrastructure products, reiterated the company’s intention to remain a software vendor.
But while VMware hasn’t been able to pull the trigger quite yet, don’t think that change isn’t coming for IT.
“Fast forward ten years. If you need a new application, the first thing you do is look for a SaaS version. If there is one, you’re done,” said Labourey. “If not, you go to PaaS, and you’ve never once touched an IaaS layer,” he said, much less dedicated infrastructure.
That’s a sea change for IT – the bulk of VMware’s customers, and what VMworld is all about. If everything is being done in PaaS and SaaS, “you’re not building IT stacks anymore,” rendering much of what IT folks do on a day to day basis irrelevant.
“It’s going to be a few years before the tsunami really hits,” Labourey said, “but if you want to stay in IT, you should go work for a cloud provider, or learn how to code.”