Cloud Data Centers:
Environmental Controls & Sensors

by Geoff Keston

Docid: 00021343

Publication Date: 1902

Report Type: TUTORIAL

Preview

Data center environmental management has grown into a
formalized process. As cloud services become more complex and more
important to enterprises, concerns about the efficiency and
effectiveness of data center management will continue to grow, creating
a greater burden for IT administrators and, increasingly, executives.

Executive Summary

[return to top of this report]

Years ago, controlling the temperature, power consumption, and
other environmental factors of a data center was considered important
but relatively easy.

Related
Faulkner Reports

Energy Efficient Computing Tutorial

Today, however, data center environmental management has grown
into a formalized process. And it is not only IT administrators who
are increasingly focusing on it. The subject is also being elevated to
the executive level in many enterprises. Cloud services have in
particular created the need for formalized management of data center
environments. As more services have been transitioned to the cloud and
as the technology for provisioning these services has become more
complex, data centers have become larger and more mission critical.

All signs suggest that this trend will not only continue but
accelerate: cloud computing is becoming even more pervasive, there are
now standards for energy management practices, and datacenter
management has become more formalized.

Description

[return to top of this report]

The growing popularity of cloud computing has forced
enterprises to manage larger data centers that deliver more complex
services, such as Big Data and mobile access. This trend has put
demands on enterprises not only to more effectively manage capacity,
but also to more efficiently control environmental conditions. The need
to improve environmental management became a concern in the industry several years ago, as noted in a report from
analyst group IDC: “[A]dding capacity without addressing underlying
issues in coordination and management will provide, at best, only
short-term gains. Adding capacity as a ‘solution’ to issues in the
datacenter is akin to putting a Band-Aid on the underlying problems: A
simplistic, brute force approach on its own won’t work.”¹

The scope of environmental concerns is broad. Heating,
ventilation, and air conditioning systems protect devices against dust
and dangerous temperatures. Clean agent fire suppression systems
extinguish fires by using substances that will not damage
electronic devices. Power conditioning equipment guards against voltage
fluctuations that can destroy data center equipment.²
Specialized coverings protect equipment against water pipe bursts,
while ground sensors warn of flooding from below. Guarding against
static electricity is also a key concern; it can be addressed in a
variety of ways, from carefully placement of equipment to treating
floors and devices with chemicals. And there are concerns that
are specific to certain circumstances – for instance, organizations in
earthquake-prone areas can also consider seismic engineering (the
design of facilities to withstand earthquakes).³

Threats to data centers can be monitored by an array of sensors, each
designed to detect a particular element such as humidity, heat, water,
or
smoke. The sensors may be integrated via a single hardware controller
that
sends alerts (e.g., through text messaging or e-mail) and that
facilitates
the Web-based configuration of danger thresholds, reports, and other
factors.

Environmental management for cloud data centers can be viewed
from two perspectives: the point of view of an organization creating a
cloud, whether public or private, or the point of view of an
organization hiring a cloud provider. Organizations that manage cloud
services must shoulder the responsibilities themselves or use a
third-party provider to assist with at least some aspects of the
process. Customers of cloud services can trust their providers to
effectively handle these demands, or they can take an active role in
verifying that providers are meeting their requirements. The latter
approach is more prudent but can be difficult if providers keep many
aspects of their data center designs confidential.

In any scenario, the job of controlling a data center’s
environment is hard. For instance, to focus on just one issue, heat,
organizations must consider how much heat is produced by all of the
data center’s equipment, how that heat is dispersed in the environment,
and how factors such as ceiling heights and room layout affect
temperature dynamics. And whereas traditional data centers grew in ways
that were easier to forecast, data centers used for cloud services are
likely to grow rapidly, even suddenly, as new customers come on board,
services are transitioned from client/server architectures, and other
trends pressure enterprises to add new capabilities.

And
environmental controls like computer-based HVAC systems
are a potential attack vector for hackers. The problem arises both
because hackers will target any system they can find but also because
some environmental systems don’t have adequate security and
organizations may not take the necessary configuration steps.
Describing this problem, Niall Browne, Chief Security Officer at
cloud-based business intelligence company Domo, says
that “[serverless offerings]
often have default passwords and have not been patched in years, as the
manufacturer was slow to release upgrades, or the customer was hesitant
to deploy them for fear of causing a service interruption to critical
functions.”⁴

Current View

[return to top
of this
report]

Over the past few years, several trends have increased the
size and the environmental needs of data centers. Industries such as
health care have promoted digital initiatives that demand intensive
computing power, and services like Big Data and mobile computing have
increased the use and complexity of data centers. In addition, data
center consolidation has become a popular strategy for improving
efficiency and reducing costs. Consolidation has become an especially
big effort within the US federal government, and the push for
greater efficiency has led to significant cost cuts.⁵And
in 2016, the US government issued an order that federal agencies cannot
build new data centers unless they can show that doing so is necessary.⁶This initiative is expected to be extended until at least 2020, but recently some changes to it have been proposed.⁷

Collectively, these trends have changed the nature of data
center management. Beginning several years ago, many organizations
started to put much greater emphasis on how they measure and control
energy use in their data centers.⁸In the past,
the demands for cooling and other controls were modest enough that
organizations could often just over-provide the resource, for instance,
simply “cranking up” the air conditioner to prevent systems from
overheating. But today, the environmental needs of a data center are
too great for this strategy because over-providing would be too
expensive and
cumbersome. Instead, it has become critical to carefully plan for and
measure environmental conditions, ensuring that controls are reliable,
affordable, and effective.

The greatly increased importance of controlling data center
environmentals has led to the development of data center infrastructure
management (DCIM) products. DCIM tools provide unified control of
factors like temperature and power usage as well as performance and
reliability functions such as asset management and capacity planning.
The leaders in this field include the following:⁹

CommScope
Cormant
FNT
Nlyte Software
Panduit
Schneider Electric
SolarWinds
Sunbird Software
Vertiv

Other large-scale data center management tools come from Manage Engine and Paessler, for example,¹⁰ and for a smaller, hardware-based tool, one option is the Black Box AlertWerks ServSensor.

Another
key development is the 2016 publication of the ANSI/ASHRAE Energy
Standard for Data Centers. According to ASHRAE, the standard
“establishes the minimum energy efficiency requirements of data centers
for design and construction, for creation of a plan for operation and
maintenance and for utilization of on-site or off-site renewable energy
resources.”¹¹

Outlook

[return to top of this report]

Over the next few years, considerations about cooling, power,
and other environmental issues will come into play in an increasing
number of IT decisions. Even decisions about what processors to use in
servers can raise environmental concerns. (For instance, Intel offers
processors that are designed to help data centers better manage power
usage and heat output. The processors monitor power needs and
automatically adjust consumption.) Interest in the subject will
increasingly grow beyond IT and reach more corporate departments. In
particular, energy considerations will be viewed as a strategic concern
and as an important financial metric. In the coming years, this focus
will increase and the field of IT energy management will grow larger.

Efforts to contain energy use appear to be working. While data
centers have proliferated and grown larger, the overall amount of power
they consume has remained about even.¹² Whereas
years ago power consumption was growing rapidly along with the boom in
data centers, recently, new designs and technology have curbed energy
demands. Their is still room for improvement, however. The advances in
energy saying techniques have mostly been implemented by the very
largest data center operators, but they have not been used widely among
other companies.¹³

Energy management will also grow more formalized, a process that has
already
begun in part because of standards such as ISO 50001, which “specifies
requirements for establishing, implementing, maintaining and improving
an
energy management system, whose purpose is to enable an organization to
follow a systematic approach in achieving continual improvement of
energy
performance, including energy efficiency, energy use and consumption.”¹⁴
Additional standards that apply include the US Environmental Protection
Agency’s (EPA) ENERGY STAR and Underwriters Laboratories’ UL2640.¹⁵Energy
concerns are also increasingly influencing more aspects of data center
design. For example, rack designs are increasingly using direct current
in
order to minimize how many power supplies are needed to support data
center
equipment.¹⁶

The management of cloud data centers will increasingly be
influenced by a
trend away from centralized models and toward the greater use of
distributed
resources. “A major shift in IT’s role in the enterprise is imminent –
or
has already happened unbeknownst to the enterprise IT professional,”
says a
study by the Uptime Institute.¹⁷ It predicts
that “IT will need
to move away from its role as a slow-moving centralized service
provider, as
IT assets become more distributed across locations and platforms, and
instead provide corporate governance across the various business lines
–
evaluating security, costs, and performance of IT for end users.”

The management of cloud data centers will increasingly be
influenced by a
trend away from centralized models and toward the greater use of
distributed
resources. “[D]ata center capacity in enterprises will not so much shrink but
fragment from large and often inefficient data centers into distributed
‘fleets’ of smaller, hardened data center ‘nodes,'”
says a
study by the Uptime Institute.¹⁸
It predicts
that some of the edge capacity may be hosted in “[m]icro-modular
data centers,” which it defines as “IT cabinets that are encapsulated
in their own protective shell with built-in cooling, connectivity,
physical security, shock absorption, and, when required,
uninterruptible power supply.” These cabinets, if put into wide use,
could change the environmental considerations of data center management.

Recommendations

[return to top
of this
report]

A good first step toward better environmental accountability
is to make its
management a formal process. Most enterprises already perform some of
the
tasks necessary for controlling a data center’s environment, such as
regulating temperature and measuring energy consumption. But these
tasks are
likely to be performed in isolation, without clear goals. Systematizing
environmental management will make formerly isolated efforts more
effective
and efficient.

This formalization may involve elevating environmental
management to a corporate-level concern. As described in a report
published by ENERGY STAR, a joint program of the EPA and the US
Department of Energy, it is wise to “[t]ake both a longer and a broader
view of investments and strategic decisions about energy.”¹⁹
The report goes on to recommend that organizations “[m]ake major
company strategic decisions (e.g., acquisitions, technology choices,
and facility location) with energy cost, use, and supply in mind.” The
development of metrics helps to formalize a process and integrate it
into corporate-wide strategic planning. In respect to power, for
example, the most popular metric is power usage effectiveness (PUE).²⁰For instance, the federal government’s 2016 Data Center
Optimization Initiative uses PUEs as a metric to define the energy
benchmarks for government agencies.²¹

A trend
that relates to environmental management is the growing use of software
for managing the performance and operation of cloud and virtual
environments. These tools help administrators configure and monitor
cloud and virtual data centers, so that they use resources more
efficiently (among other goals). This bears directly on environmental
management because more efficient usage of resources can reduce power
consumption. Leading vendors providing such software include Cisco, HP,
IBM, Microsoft, ServiceNow, and VMWare.²²Considering
environmental management in the context of overall cloud and virtual
data center management can help enterprises to better achieve the goals
of both practices.

References

¹Richard L. Villars. “The
Datacenter’s Role in Delivering Business Innovation: Using DCIM to
Enable a Common Management Approach.” IDC.
November 2012.
²Scott Lowe. “Keep Your Power
Clean with the Right Conditioner.” TechRepublic.
May 17, 2002.
³ Stanford University. “Seismic
Engineering Guidelines.” Stanford University.
October, 2017.
⁴ Maria Korolov. “Data Center Infrastructure, the Often-Overlooked Security Risk.” Data Center Knowledge. April 17, 2018.
⁵ Patrick Thibodeau. “Republicans
and Democrats Back Federal Efforts to Scale Back U.S. Data Centers.” Computerworld.
September 26, 2014.
⁶Yevgeniy Sverdlik. “White House
Orders Federal Data Center Construction Freeze.” Data Center
Knowledge. March 4, 2016.
⁷Tajha Chappellet-Lanier. “OMB Proposes Changes to Data Center Optimization Policy.” FedScoop. November 26, 2018.
⁸ Clemens Pfeiffer. “Twelve Data
Center Energy Management Trends for 2012 and Beyond.” GreenBiz.com.
January 25, 2012.
⁹ April Adams, David J.
Cappuccio, Neha Kumar, and Tiny Haynes. “Magic Quadrant for Data Center
Infrastructure Management Tools.” Gartner.
October 10, 2016.
¹⁰ “Reviews for IT Infrastructure Monitoring Tools.”
Gartner.
¹¹ “Data Center Standard Published By ASHRAE.” ASHRAE. September 13, 2016.
¹² Katie Fehrenbacher. “Data
Centers Are No Longer the Energy Hogs they Once Were.” Fortune.
June 27, 2016.
¹³ Ibid.
¹⁴ “Energy Management Systems (ISO
50001: 2011).” International Organization for Standardization.
¹⁵ Clemens Pfeiffer. “Ten Key Data
Center Energy Management Trends for 2013.” Energy Manager
Today. January 24, 2013.
¹⁶ Stephen J. Bigelow. “Ten Data
Center Trends Driving Change in 2015.” TechTarget.
June 18, 2015.
¹⁷ Matt Stansberry. “Uptime
Institute Global Data Center Survey.” Uptime
Institute. 2018.
¹⁸ Claire Swedberg. “Central
Intelligence: The IoT Brings Big Changes to Data Centers.” Electrical
Contractor. October 2016.
¹⁹“The Future of Corporate Energy
Management.” Energy Star.
²⁰Clemens Pfeiffer. “Must-Know
Metrics for Data Center Energy Management.” Environmental
Leader. April 3, 2012.
²¹Frank Konkel. “Here’s the
Government’s New Policy for Optimizing Data Centers.” Nextgov.
March 1, 2016.
²² Dennis Smith and Milind
Govekar. “Market Guide for Cloud
Management Platforms.” Gartner. April 25, 2017.

Web Links

[return to top
of this
report]

Black Box: https://www.blackbox.com/
Cisco: http://www.cisco.com/
CommScope: http://www.commscope.com/
Cormant: http://cormant.com/
Energy Star: http://www.energystar.gov/
FNT: https://www.fntsoftware.com/
HP: http://www.hp.com/
IBM: http://www.ibm.com/
Intel: http://www.intel.com/
ISO 50001:2011: http://www.iso.org/iso/catalogue_detail?csnumber=51297
Manage Engine: https://www.manageengine.com/
Microsoft: http://www.microsoft.com/
Nlyte Software: http://www.nlyte.com/
Panduit: http://www.panduit.com/
Paessler: https://www.paessler.com
Schneider Electric: http://www.schneider-electric.com/
ServiceNow: http://www.servicenow.com/
Sunbird Software: http://www.sunbirddcim.com/
UL: http://ul.com/
Vertiv: https://www.vertivco.com/
VMWare: http://www.vmware.com/

About the Author

[return to top
of this
report]

Geoff Keston is the author of
more than 250 articles that help organizations find opportunities in
business trends and technology. He also works directly with clients to
develop communications strategies that improve processes and customer
relationships. Mr. Keston has worked as a project manager for a major
technology consulting and services company and is a Microsoft Certified
Systems Engineer and a Certified Novell Administrator.

[return to top of this report]