In June 2015, I was on a panel of bloggers who discussed OpenStack with HP’s Converged Cloud group. The team was very forthcoming about HP’s Helion OpenStack efforts and their customers’ adoption of Helion OpenStack. I walked away with a better understanding of how enterprise customers are viewing Helion OpenStack, and potential use cases for the platform.
My initial thoughts about OpenStack
When OpenStack was announced by Rackspace and NASA five years ago, I was uncertain about the initial use case. I discovered the challenge wasn’t an issue of understanding OpenStack but rather understanding the use case for cloud in the enterprise.
At that time, the concept of virtual machine (VM) centric cloud infrastructure was prevalent. In previous enterprise solutions such as VMware vCloud, the consumer of infrastructure resources would interact with a web-based catalog, which offered various VM configurations.
OpenStack managed resources are consumed using API requests; the consumption model is similar to Amazon Web Services (AWS) in that infrastructure management was moved higher up the stack to the development team. There are challenges with applying this model to the enterprise, which is why OpenStack was always a hard sell to traditional infrastructure managers.
The OpenStack community initially played up interoperability between OpenStack environments. In theory, a customer could stand up on-premises OpenStack-based infrastructure that could expand to a wide variety of OpenStack public clouds; the result would be a public cloud ecosystem that could compete with the scale of AWS.
The theory of an OpenStack managed infrastructure is one thing, but how have organizations deployed and used OpenStack?
HP Helion OpenStack deployments
OpenStack has proven to be a solid platform to build a cloud-native application. Speaking to HP, the primary use case has been departments looking to build a new cloud-native API. Customers have chosen to create dedicated infrastructures separate from their traditional infrastructure.
HP customers focused on tuning the infrastructure for the single application use case. However, customers can use the initial deployment as a foundation for future, expanded Helion OpenStack infrastructures capable of supporting multiple cloud-native applications. According to HP representatives, most of HP’s private cloud customers are in the initial phases of infrastructure buildouts.
HP would like customers to leverage HP Converged Infrastructure hardware to deploy pods to run Helion OpenStack. Single pod deployments of OpenStack slowly introduce a new technology and a new IT delivery model. Operations teams get a chance to understand the impact of OpenStack, and developers start to understand how to build cloud-native applications.
The end-to-end vision is to provide an hybrid cloud solution. HP’s preference would be to leverage HP hardware on-site and a compatible OpenStack public cloud such as HP’s Helion Public Cloud solution.
In future articles, I’ll explore ways to integrate private cloud infrastructures and DevOps management models into existing infrastructures.
What about your organization?
What is your strategy for integrating private cloud? I’d love to hear your thoughts.
Note: TechRepublic and ZDNet are CBS Interactive properties.