With little fanfare and a few months' warning, HP is pulling the plug on its Helion Public Cloud platform. HP says it'll "double-down on our private and managed cloud capabilities" by way of Helion CloudSystem and Helion OpenStack.
For a company that was vocal about making a public cloud with enterprise appeal, the move is a major concession of the public-cloud field to its established players. But it's also a sign HP has wised up about the real opportunities in building clouds for enterprises.
Here are the five most significant insights we've gleaned from HP canning Helion Public Cloud and turning to a hybrid-cloud product instead.
1. HP's public cloud never had a chance
In truth, it would have been brutally hard for HP to make a public cloud that competed in a meaningful way. By the time HP got into the game, the public cloud was already split between Amazon, Google, and Microsoft. Contend with the trio at your own peril.
Amazon, for one, is more than a default go-to choice; it's also a model for others to follow. (How many other cloud products and how many other clouds offer API compatibility with AWS as a selling point?) Google and Microsoft may not lead in numbers, but they've chiseled niches for themselves. Google has shown savvy leveraging of open source and open standards, while Microsoft leans on its existing presence in enterprises everywhere.
HP, on the other hand, faced an uphill fight from the beginning. It came late to an already mature market, offered little in terms of unmistakable advantage, and its split into two companies -- one enterprise-focused, the other consumer -- didn't afford it any new benefits.
You can't say HP's public cloud wasn't well-engineered or didn't come from a company with private cloud services that were well regarded by its customers. However, none of those boons translated into substantial market share or mind share.
2. Promises of no lock-in aren't enough for enterprise cloud providers
HP built its public cloud with OpenStack because it didn't have to reinvent the wheel. However, the promises of "no lock-in" that typically go with OpenStack didn't seem to matter much.
Though perhaps proprietary, Amazon's cloud is also thoroughly documented and well understood -- those who choose it often reap rewards far exceeding the inconvenience of lock-in. If HP hoped the freedom of movement provided by open source would be appealing, it forgot that the same tactic alone accomplishes little. The desktop Linux world had been trying the same argument on users for years (open source! no lock-in! software freedom!) to get them to abandon Windows, with little success.
HP wanted to stand apart by presenting enterprise developers with a complete enterprise app lifecycle solution, where there's little worry about moving from development to production. It was a smart move: Lock-in tends to be a more abstract problem, but dev and deployment issues are an immediate, in-your-face pain. Maybe it'll pay off more now that the distraction of engineering and promoting a public cloud has been swept away.
3. It won't be about leveraging OpenStack alone
OpenStack tends to dominate the conversation about private and hybrid cloud, but it's far from the whole picture, especially when it comes to what HP plans to offer.
For starters, there's Eucalyptus, bought by HP for its ability to provide AWS-compatible private or hybrid cloud architecture. Using Eucalyptus to build bridges to and from Amazon is a stated part of HP's revamped hybrid plan. Likewise, HP plans to keep offering Cloud Foundry for those wanting a PaaS in a variety of environments.
HP has done a fair amount of work integrating OpenStack and Cloud Foundry, so the two are complementary, not exclusionary. Still, there's little question OpenStack will continue to be both the substrate and the substance of HP's cloud strategy.
4. Private and hybrid enterprise clouds are more HP's concerns
When HP has received accolades and recognition for any of its cloud work, it's been for its private and hybrid cloud offerings, not its public product.
Forrester Research found in late 2013 that HP had been valued consistently highly for what it offered customers wanting to build a private cloud; of late, Forrester gave the nod to HP as a leading provider of private cloud solutions in China. Beyond that, HP had already made purchases and technology decisions better suited to hybrid infrastructure, like the above-mentioned Eucalyptus.
HP also has traditionally enjoyed good reach with enterprise customers due to existing marketing and sales apparatus. It's a go-to name and has stumped for better individual service and more competitive SLAs, a perpetual vigilance point for enterprises.
5. HP's right: Hybrid infrastructure probably is the future of enterprise IT
HP is wise to proclaim hybrid as the default stance for enterprise clouds. Not everyone can, or should, go full-bore public with their infrastructure (regulatory concerns, logistics), but keeping everything behind the firewall cheats enterprises out of some of the best reasons to have a cloud in the first place (elasticity, cutting hardware costs).
HP has a plan to make it a reality, but it'll face two major challenges moving forward. First, it has to find a way to make its solutions compelling to an audience that isn't merely existing HP customers. Amazon may be more lackadaisical about appealing directly to the enterprise, but that also means nearly anyone can be an AWS user.
Second, and most critical, HP has to not let other ongoing upheavals get in the way. The enterprise side of HP was by all accounts highly dysfunctional when current CEO Meg Whitman took the reins, so merely splitting the company doesn't mean its enterprise side will magically deliver.