A provocative new study from the UK's BroadGroup argues that investment in private clouds today will likely result in an increase in public cloud utilization down the road. It comes down to a matter of trust -- first in the cloud itself, then in the cloud maintained by someone else. Will this ultimately lead to critical applications on the cloud? Perhaps, as the firm's Marion Howard-Healy explains.
Arthur Cole spoke with Marion Howard-Healy, senior consultant, BroadGroup.
Cole: Your latest report on private clouds predicts a fairly healthy adoption rate over the next few years. But since these are still internal structures, how exactly will this boost data center outsourcing as well?
Howard-Healy: So in Europe, we already see the first stage of data center outsourcing of the private cloud. Established hosted managed service providers are developing strategies to move customers up their service value chain. The current focus is on hosting internal cloud services, but over time as security, SLA and other business issues are addressed, one can foresee a scenario whereby customers will become very comfortable in buying more of their internal cloud from their hosted managed service provider with whom they have trusted relationships. These established providers include telecom service providers, data centers and other host providers. Organizations will be doing less systems themselves, and external services will make up the shortfall.
"The cloud challenges business as usual because it impacts designs and architectures, funding models, provisioning and procurement, service response levels and SLAs, as well as people, policies and business processes."
Cole: As IT-as-a-service offerings grow, how will they change enterprise business and operational models?
Howard-Healy: The cloud challenges business as usual because it impacts designs and architectures, funding models, provisioning and procurement, service response levels and SLAs, as well as people, policies and business processes. Cost reduction is often the initial motivator for using the cloud, but understanding total cost of ownership and the benefits gained from greater business agility should be top of mind.
The on-demand, self-service aspect of the cloud introduces a new operating model for users because it brings together infrastructure and business rules and processes. This is a profound change for enterprises because it aligns what IT produces with what the business consumes and leads to much closer working between IT and the business units. Automation and access to resources using APIs replaces manual intervention. The flip side of self-service is that service providers -- including in-house data center operators -- must be able to meter and bill the different types of resources -- intellectual, electrical and physical -- in the cloud.
The cloud will change behaviors in both the enterprise and data center markets. It introduces multi-contractual regimes, increases collaboration among supply partners, and fosters a need for global standards. Questions on legislation, standards, SLAs and security are key issues to resolve near term.
Cole: Are we still talking about mostly non-critical applications and data, though? Will hybrid or public clouds ever gain a toe-hold in critical areas?
Howard-Healy: Discussion is evolving around core-vs.-context in terms of applications that can be moved to the cloud. Generally speaking, core/critical apps will not move to the public or hybrid cloud in the near term, although there may be some exceptions. Today we are talking mostly of non-critical applications, as you say. However, if issues of security, SLAs, standardization and interoperability are addressed by vendors to business user satisfaction -- sufficiently to allay current user concerns -- there is no reason for hybrid and public clouds to gain a toe-hold in critical areas -- but that is in the longer term, 10 years.