In Romeo and Juliet, Shakespeare asks, “What’s in a name?” Juliet goes on to say, “That which we call a rose / By any other name would smell as sweet.”
Simply stated, a word alone does not define a rose; we know it by its attributes.
However, the same isn’t true in government IT. A data center by any other name might be a computer room or a server closet. It all depends on which OMB data center optimization memo you are reading.
Recently, as a part of the Cloud Smart strategy, OMB released a draft of a new memo on the Data Center Optimization Initiative (DCOI). The memo seeks to maintain the overall goal of closing or consolidating data centers, based on a new definition. There are a number of new metrics in the memo around availability, virtualization, and power efficiency. These are important, but in the end, the only metric that really counts is mission delivery.
To put these changes in context and understand the mission impact, let’s look back to where it all started.
The government’s first attempt to consolidate data centers was the Federal Data Center Consolidation Initiative (FDCCI). Started in 2010, the goal was to close 40% of the government’s data centers and save $5 billion by 2015.
The policy organized data centers into core or non-core. A big criticism of FDCCI was that it defined a non-core data center to include small server closets and self-contained computing facilities, increasing the number of non-core data centers from an initial estimate of 3000 to over 7100. However, in a classic case of what gets measured gets done, CIOs spent most of their time consolidating these small spaces, yielding only minimal savings.
The passage of the Federal IT Acquisition Reform Act (FITARA) in 2014 brought new scrutiny on data center consolidation and adoption of cloud computing technologies. This led OMB to revisit FDCCI. In 2016, the Data Center Optimization Initiative (DCOI) was born.
Wisely, DCIO was not just about consolidation. It also focused on strategies to move applications to more efficient infrastructures, with a focus on energy savings. Further, it reclassified data centers as tiered and non-tiered, reflecting industry trends. However, non-tiered data centers included “rooms with at least one server, providing services”. This meant that agencies needed to classify print servers at their remote offices as a data centers, skewing counts.
As of August 2016, the agencies collectively identified 9,995 data centers under this definition. According to GAO, the savings from closing these small facilities was far less than expected.
The FITARA Enhancement Act of 2017 extended the data center requirements of FITARA until October 1, 2020, setting the stage for a revised approach to data center optimization.
The new OMB memo builds on past efforts and refines the government’s focus. The memo acknowledges that agencies have “seen little real savings from the consolidation of non-tiered facilities, small server closets, telecom closets, individual print and file servers, and single computers acting as servers.” Consequently, agencies will no longer have to consolidate these computing resources, allowing CIOs to focus their efforts on the consolidation of full-fledged data centers. All will welcome this long overdue change.
The memo emphasizes risk-based decision-making and service delivery as key considerations in evaluating cloud technologies and data centers. I applaud the memo’s acknowledgment that commercial cloud services are “not always more cost-effective alternatives to agency-hosted services”. This was my experience as a CIO. Cloud migrations are complex and require careful consideration to ensure they are the best choice. You must look before you leap to the cloud.
Given data egress costs from the cloud, it’s critically important for agencies to be able to move applications and data between environments. The next version of the DCOI memo should address migrations between environments, including commercial clouds and agency managed infrastructures to create agility and avoid vendor lock in.
As noted in the memo, new technologies, like mobile and “Internet of Things”, are rapidly growing agency data volumes. Moreover, it’s necessary to store, govern, protect, and secure all of this new data. These activities make it possible for data to be available when needed. The lack of attention to these data management core competencies delays projects, drives up the cost, and complicates migrations.
The memo also calls for the creation of a DCOI Strategic Plan. As part of this document, agencies should be required to detail an Enterprise Data Governance Strategy for their data across all data center and cloud environments.
Only after you have taken steps to understand what data exists, how it is classified and protected, who owns it and who has access to it, can you make a fair determination of how (or if) to make that data available or where to migrate it. The quicker you can conduct these data inventories, the more effective your data center consolidation activities will be.
Further, these activities support automated, policy driven decisions for moving data to the appropriate type of storage solution, based on the type of data. This drives cost reductions through storage optimization and created opportunites for AI/ML technologies to monitor system health and detect potential issues as called for in the Automated Infrastructure Management section of the memo.
In the end, the success of new DCOI policy will be judged not on the number of data centers the government closes, but how it supports mission enablers like the Cloud Smart and the Federal Data Strategy.
The mission always runs on data. So, it’s essential that data management be an integral part of these initiatives, no matter what name you give them.
This piece was originally published via GovLoop.com on December 20th, 2018.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.