Government computing resources, like any other government procurement, used to be purchased by agencies for those agencies... and nobody else. It didn't make sense to share, because the very concept of sharing compute power didn't even exist. Now in an almost unprecedented shift of philosophy, the U.S. Government is one of the world's leading adopters of private cloud infrastructure. In order to slash costs fast, it's moving to the cloud sooner than almost anyone else.
Now, some government agencies, departments, bureaus, and divisions that no longer have any reason to avoid maintaining their compute resources separately from one another, also have no viable reason for staying separate from one another. IDC analyst Shawn P. McCarthy has discovered, and discusses in a newly published report, local governments are rapidly slashing costs by purchasing compute power and capacity on a metered basis from state governments.
Sponsor
Regional federal offices are doing the same by purchasing capacity and power from agencies upstream. The result is the creation of multi-agency, regional hubs procured for the first time not by any one division or department, but through new and emerging consolidation efforts that are bringing nearby local services and regional departments together for the first time.
McCarthy sees three pillars of private government cloud services, depicted in the diagram above. "Type 1" are analogous to private clouds in private enterprises, hosted and managed entirely in-house. "Type 2" are maintained by vendors or consultancies while hosted in-house. "Type 3" are maintained by vendors and hosted off-site. Now, you may be wondering, just how can you have a private cloud that's off-site? McCarthy explains that hardware for government private clouds are always devoted 100% to government services, even when they're located elsewhere.
The most trusted compute service provider that a local government or department may find that complies with state and federal requirements, naturally, is a state or federal provider, McCarthy writes. There are upshots on both sides of this relationship: Compliance with state and federal data privacy and reporting regulations is more strongly ensured. Also, when several smaller agencies pool together to purchase computing power in volume, that volume helps drive down costs for each of them.
In turn, the bigger agencies recoup some of their costs for having procured upgraded, 21st century hardware.
"As government workers started interacting with systems located miles away from their offices (rather than down the hall), both CIOs and CFOs have seen that their organizations no longer need to be in the business of owning and operating multiple types of IT solutions," McCarthy writes. "Today, common functions, such as e-mail, human resource management, procurement, and other business functions, can be obtained from cloud providers at a highly competitive price and with increasingly acceptable service-level agreements. Once they do the math, cloud is often the direction these organizations choose to move."
But although McCarthy's report is subtitled "The New 'Trickle Down' Effect That's Boosting State and Local Computing," he takes note of one quite unexpected, though not unforeseeable, phenomenon: Smaller departments and even state government offices could find themselves recouping costs by outsourcing cloud computing power upstream to bigger agencies. He writes, "Any level of government can, in theory, offer services to any other government office. It's just a matter of building the infrastructure and having a way to offer reliable cloud-based services to government clients."
McCarthy foresees a scenario where the complete upheaval in government hardware architecture could soon revolutionize the way its applications are created, sold, and delivered. For example, geological survey software may only be important to a few departments, and for that reason alone they tend to be expensive. Once it moves to a service model, conceivably a larger government agency (hypothetically, the U.S. Dept. of the Interior) could deliver "survey-as-a-service," if you will, to smaller agencies that would pay not for the software but for the actual tasks performed. This could be the next wave of government structural reform that both reduces costs while improving efficiency.
Discuss
0 comments:
Post a Comment