OpenStack is many things. It’s an open-source cloud framework launched initially to challenge VMware and Amazon Web Services. It’s a set of compute, networking and storage components and associated tools. And its very definition is evolving along with the underlying code. That fluidity is one reason big production deployments are still in short supply, according to an analyst and press panel at this week’s OpenStack Summit in Paris.
Perhaps the main hurdle to wider adoption remains the sheer difficulty of deploying OpenStack. Panelist Al Sadowski, an analyst with 451 Research, likened OpenStack to old-fashioned oatmeal.
“In the old days of oatmeal your mom had cardboard canister and water and a pot and if you didn’t put in enough water it was lumpy and it was a pain to clean the pot. OpenStack needs to [become ] instant oatmeal. Everything’s needs to be in there, you just rip off the top and add water,” said Sadowski. Not only that — it needs to be the variety pack of instant oatmeal. “So there’s a big-data version, a private cloud for web hosting version. It needs to be that simple to consume to have wider adoption.”
The panelists largely agreed that the OpenStack brand is powerful. But for non-technical consumers of cloud that may not matter.
“Do enterprises really care about OpenStack or are they just thinking about private cloud or managed private cloud and don’t care what’s underneath … if it’s OpenStack or [company]VMware[/company] or [company]Microsoft[/company] –as long as the use cases can be developed,” said René Büst, senior analyst at Crisp Research.
What struck me about much of this discussion was how similar it is to talks I’ve heard at various OpenStack events over the four years of its existence. But then again, cloud is hard and four years isn’t that long a time.
There are a lot of other good insights in here so check out the video of the session linked below (thanks to Büst for the pointer.)