OpenStack Summit Emphasizes Emerging Deployment Models
The OpenStack Summit kicks off with discussion on moving beyond just the private cloud model to a managed private-cloud-as-a-service approach
he OpenStack Summit kicked off here today with multiple announcements and an emphasis on the evolution of the cloud deployment model.
Jonathan Bryce, executive director of the OpenStack Foundation, said during his keynote that there has been a 44 percent year-over-year increase in the volume of OpenStack deployments, with OpenStack now running on more than 5 million compute cores around the world.
Although OpenStack has had success, the path has not been a straight line upward since NASA and Rackspace first started the project in June 2010.
“We’re now at a major inflection point in the cloud,” Bryce said.
Openstack Summit Boston
OpenStack is now moving to the second generation of private cloud, according to Bryce. He noted that the first generation of private cloud was all about hyper-scale cloud deployments from large vendors like eBay and Yahoo that are now running OpenStack at massive scale. In contrast, the second generation of OpenStack private clouds is about broader use cases and multicloud deployment models.
“What was interesting is that as the community grew and people started contributing to OpenStack, it went into a new set of arenas that the public cloud didn’t handle,” Bryce said.
Among the items that OpenStack was quick to embrace is IPv6 addressing support as well as network functions virtualization (NFV).
The first generation of private cloud was concerned with issues around technology and IT staff talent, according to Bryce. In contrast, the second generation of private cloud deployments is more concerned about culture and process.
Back in 2010, the line between private and public cloud was also easily understood, with organizations often choosing one or the other. Bryce said cloud environments today are diverse with companies realizing they need a range of options, including both internal cloud and public cloud. There are three C’s—compliance, cost and capabilities—that typically help determine where a cloud workload should land, he said.
Bryce added that OpenStack today works well also as a private hosted private cloud. There is also an emerging area of OpenStack deployment mode called the remotely managed private cloud, or private-cloud-as-a-service.
To that end, the OpenStack Foundation today announced a new category in its online marketplace for vendors. The new Managed Private Cloud offering category provides a listing of all the vendors that offer that service, which includes Mirantis among other vendors.
Mirantis co-founder Boris Renski used his time on the keynote stage to detail some of the mistakes of the private cloud.
“The term ‘private cloud’ is a toxic word in the enterprise,” Renski said. “The reason why is because private cloud evolved to become a marketing term that unfortunately from time to time is associated with failure.”
In his view, the reason why private cloud sometimes fails it because it doesn’t always deliver the same benefits and experience that have been promoted by public cloud vendors. Those benefits include cost, integration and ease of use.
“We want to have managed open cloud as a service that focuses on the right problems,” Renski said.
Originally published on eWeek