A draft of the new Data Center Optimization Initiative (DCOI) policy released Monday reorients the focus of DCOI toward optimization–rather than data center consolidations and closures–and revises metrics for the program.
The draft memo from Federal CIO Suzette Kent, which has been released for public comment until December 26, would replace guidance from 2016, and would align DCOI with the new Cloud Smart policy from the Office of Management and Budget (OMB).
Among changes in metrics, the draft guidance would: remove energy efficiency and facility utilization as metrics; revise metrics for virtualization, energy metering, and server utilization; and add a metric for availability.
“Much of the ‘low-hanging’ fruit of easily consolidated infrastructure has been picked, and to realize further efficiencies will require continued investment to address the more complex areas where savings is achievable. While optimization will be the new priority, consolidation and closures should continue wherever applicable. OMB will focus on targeted improvements in key areas where agencies can make meaningful improvements and achieve further cost savings,” the draft guidance states.
In addition, the guidance places priority on virtualization as a metric, followed by availability, energy metering, and server utilization, in that order.
“The ability to assign resources dynamically based on need is key to delivering timely services, and is an important concept in the Cloud Smart strategy. As such, OMB prioritizes the increased virtualization of Federal systems as critical for IT modernization efforts, to drive efficiency and application portability,” the draft guidance states. Agencies will be able to report virtualized systems under their cloud investments under the new guidance.
Also facing revision are metrics around energy metering and server utilization. Automated metering requirements would be loosened under the new policy, allowing agencies to request exemptions where it would be too costly or where the data center is scheduled to be closed. Agencies would also have greater leeway on server utilization, removing automated metering from the metric and allowing agencies to determine their own methodology, using CPU usage and storage space as a baseline.