Datacenter Consolidation & Optimization

Consolidation & Optimization

Data center optimization and consolidation is something that is often on the mind of those involved with the operations of large datacenters. The larger the deployment footprint, the larger the potential savings. The United States Federal Government kicked off a large data center consolidation towards the end of last year by issuing the DCOI Memorandum, and considering not many deployments of this size are in existence, it’s an interesting case, and worth a look.

Datacenter Closing

We’ll take a quick look at the goals of the consolidation, which outline a standard, high level framework that almost anyone with a large data center footprint can use to think about their own consolidation. We’ll then look at how Device42 can help realize some of the goals laid out by the DCOI initiative.

The Federal DCOI (Data Center Optimization Initiative)

The DCOI is a government program that was issued August 1, 2016 as a follow up to the DFCCI (Federal Data Center Consolidation Initiative). The goals of both initiatives are similar- Consolidation, Optimization, and Cost Savings:

Consolidate: Reduce the real estate footprint of Federal Government Data Centers
Optimize Power Usage: Reduce the overall energy usage of Federal Data Centers
Optimize Hardware Spend: Reduce the cost of hardware
Optimize Software Spend: Reduce the cost of data center software
Boost Efficiency: Reduce operational costs [by 25% by the end of 2018]
Enhance security as a side effect of doing all of the above

Some of the core strategies that are being used to accomplish DCOI’s consolidations and optimizations are detailed in the memorandum M-16-19:

A spending and budgeting freeze towards initiation of new data centers or existing data center expansions
A cloud first policy: transition to cloud data centers wherever possible, taking advantage of provisioned services such as SaaS, PaaS, or IaaS
Consolidate and/or close existing data centers by:
- Virtualize wherever possible
- Pool storage, network, and compute resources
- Eliminating duplication and sharing services where possible across agencies
- Migrating compute loads to more modern, optimized federal data centers

Measuring Progress

measuring progress

Some of the important metrics that will be used to track and measure the success of the consolidation effort are as follows:

Energy Usage – Overall Power Consumption
Server Utilization Levels – Virtualization Effectiveness
Facility Utilization – Effective use of real estate capacity, right down to the racks

These high level goals are rather universal, and the degree of each one’s importance to your bottom line will vary depending on your specific deployment and facilities.

How Device42 Can Help

It is very difficult to embark on a consolidation if you don’t first have a clear picture of exactly what you have, where, and how it’s all connected. Part of the requirements of the DCOI are the installation of automated energy monitoring tools that can report overall power usage, as well as the installation of automated Data Center Infrastructure Management tools to replace manual methods of collecting and reporting on data. Figuring out what you have, and installing tooling to benchmark where you are beginning are typical first steps in any consolidation, and are critical.

“Agencies shall install automated energy metering tools and shall use these to collect and report energy usage data in their data centers to OMB. The March 19, 2015, Executive Order 13693, “Planning for Federal Sustainability in the Next Decade,” requires agencies to install and monitor advanced energy meters in data centers by September 30, 2018”

“Agencies shall install automated energy metering tools and shall use these to collect and report energy usage data in their data centers to OMB. The March 19, 2015, Executive Order 13693, “Planning for Federal Sustainability in the Next Decade,” requires agencies to install and monitor advanced energy meters in data centers by September 30, 2018.19”

Device42’s powerful, comprehensive auto discovery can find all of the devices running on each network, populating its CMDB and producing an accurate inventory of the devices themselves, the software and services running on them, the underlying, supporting infrastructure, and their underlying interdependencies. Device42 can also autodiscover power equipment like iPDU’s and UPSs, and can monitor and/or control them, along with a variety of environmental sensors.

When it comes time to begin to look at and optimize facilities, Device42 can track Buildings, Rooms within them, the racks within those rooms, and the equipment in the racks (and even how it’s all wired, for both network & power). Color coded heat-maps let you see, at a glance, rack utilization levels with overlaid power usage bars. Combined with Device42’s flexible reporting, identifying underutilized rack space and ensuring efficient usage is a clear-cut process.

Conclusion

The underlying goals of the United States Governments DCOI consolidation fall right in line with the goals all data center operators share: Consolidation for optimal workload distribution, and managing capacity for the most efficient use of limited space and resources. Though the deployment footprint of the United States Federal Government is likely much larger than any most of us will ever work with, it’s interesting to note that the same processes and techniques apply, albeit at a much larger scale. The end goal for nearly all consolidation and optimization efforts is also the same: namely, reduce the bottom line, leaving more money in the pockets of the operators.

With the right tools, undertaking an optimization, though a challenging undertaking, can be a no brainer. Why spend more money than required for the same result? The DCOI provides a framework for thinking about how you can consolidate your data center footprints, and how you can continue to optimize as you go. With a few simple rules like all new servers are virtual, and cloud-first thinking wherever possible, you’ll be well on your way to running a more efficient operation.