OpenStack needs to get a move on it, a leading Forrester analyst says. Member companies need to start seeing a return on their investments or else they may become disinterested.
The OpenStack cloud computing project needs to get a move on it, says Forrester analyst James Staten. Specifically, he says member organizations need to start seeing a return on their investments, or else may become disinterested.
More than two and a half years old now, the OpenStack project has gained significant momentum in the past year, in large part because big-name companies such as VMware, Red Hat and IBM have joined the likes of Rackspace, Cisco, Dell and HP to contribute more than $10 million to the project.
Despite the investments, Staten points out there have been relatively few "enterprise-ready" OpenStack-powered products on the market. Canonical, SUSE, Rackspace and Morphlabs are some of the companies with OpenStack distributions so far, he notes. Meanwhile, Rackspace and HP are two of the biggest companies to use the OpenStack code to power their own clouds.
Competing open source cloud management platforms CloudStack and Eucalyptus have a "big lead" in terms of production deployments of their systems and their ability to generate revenue, Staten says. Those are "key motivations" for keeping a vendor engaged in an open source project. "OpenStack needs to start producing revenue for its partners," Staten says. "The clock is ticking."
DISASTER RECOVERY IN THE CLOUD: Vendors jump in, enterprises wade
OpenStack backers claim Staten is spreading FUD (fear, uncertainty, doubt) about the project. "It's still very early," says Jim Curry, general manager of Rackspace's private cloud business. "We absolutely want people to make money on this," which is why vendors are contributing code to the project to ensure their products and services work in the OpenStack ecosystem.
Big-name member companies like IBM, Dell, HP and Red Hat are all in the "early stages" of developing their OpenStack strategies. Red Hat, for example, just joined the project four months ago and announced a free version of its OpenStack distribution, with plans to roll out updates next year. "Big companies take months to launch products. They do it in quarters, not weeks. It will take some time," Curry says. "The momentum is there."
Staten says OpenStack can be successful, but he says the readiness of the platform is still an open question. The biggest signs of its maturity will be more vendors shipping OpenStack-powered production-ready solutions and enterprises adopting those solutions. That has been slow, so far.
Staten is not the first cloud computing analyst to offer sobering remarks about the OpenStack project. Earlier this month Gartner analyst Lydia Leong, another well-respected voice on cloud topics, warned users to beware of "dangerous myths" that are emerging about OpenStack. She says customers she's spoken with have inappropriately assumed that OpenStack inherently allows for interoperability with other clouds, which she says it may not. She adds that overall project "difficulties" are causing some members to internally reconsider their OpenStack strategies.
Krishnan Subramanian, a cloud industry blogger and analyst at Rishidot research, disagrees with both Staten and Leong, though acknowledges enterprise adoption of OpenStack has been slow. "I would give 2 more years before I write my obituary for OpenStack," he says. Compared to Google, Amazon Web Services and VMware, OpenStack member companies are spending comparatively minimal amounts of money on the project itself. These companies will make money not just based on distributing the code, but by offering products and services in support of the core code. "Expecting big money on OpenStack itself is a wrong expectation," Subramanian says.
The bigger concern, he says, is around having solid code within OpenStack that can be used by these companies to launch products and services, and by enterprises to launch their own clouds. "With Rackspace and HP running closer to OpenStack trunk -- they lag by just 2 weeks -- there is enough evidence to believe that the code is getting mature," he says. "They still need to fix a few kinks here and there and once it is done, it is just a matter of time before they could convince enterprises to trust OpenStack for their production needs. I think they have this leverage in the coming year or two."