Essential Guidance on DCIM for

Edge Computing Infrastructure

White Paper 281

Revision 0

by Patrick Donovan

Executive summary
The lack of staff or “lights out” nature of many local IT and
mobile edge computing (MEC) sites makes operations &
maintenance a challenge. This struggle worsens as the
number of sites increase. How do you maintain IT resiliency
in a cost-effective way under these conditions? It is not
practical to staff each location with trained personnel. The
answer lies, in large part, on data center infrastructure man-
agement (DCIM) software. In this paper we describe essen-
tial DCIM functions for small, unmanned edge computing
sites and attributes of next-generation DCIM solutions best
optimized for that type of environment. W e also provide
practical advice on how to get started with DCIM to better
ensure its value is realized.

RATE THIS PAPER 

 Schneider Electric – Data Center Science Center White Paper 281 Rev 0 2

Introduction Smaller, local edge computing sites - typically 1 to 4 IT racks in size - are often geo-
graphically dispersed, multiple in number, and lacking IT staff. This creates infra-
structure management and maintenance challenges (Figure 1) that make it difficult
to maintain availability of the IT in an efficient manner. Data center infrastructure
management (DCIM) software tools are critical to addressing these challenges.
Modern, next-generation DCIM apps simplify infrastructure management while im-
proving resiliency. This paper describes:

• Essential DCIM functions for ensuring high resiliency at the edge

• Attributes of next-generation DCIM platforms optimized for use with multiple,
highly distributed, and unmanned local edge computing sites
• Tips on how to get started with and use the tools

For a full, detailed explanation of what DCIM is and what all its functions are, see
White Paper 104, Classification of Data Center Infrastructure Management (DCIM)
Tools.

When & how do I

replace batteries?
Who is accessing
my equipment?
Figure 1 I need visibility
Example challenges that over all regions
arise when managing infra-
structure systems across
multiple, geographically-dis- So many alerts
persed, unmanned local and I don’t know
edge computing sites what to act on

I only want visibility to I lack expertise on

sites in MY region infrastructure equipment

Essential DCIM White Paper 277, Solving Edge Computing Infrastructure Challenges, makes the
case for embracing a collaborative ecosystem of partners and vendors, along with
functions for adopting integrated micro data center solutions as the best approach for efficiently
ensuring high maintaining IT resiliency at the edge. A critical element of the micro data center so-
lution is the next-generation DCIM system. These software apps remotely monitor
resiliency at the and optimize power, cooling, and security infrastructure equipment. Fundamentally,
edge these new cloud-based DCIM tools offer easier management from afar, big data an-
alytics and artificial intelligence technologies, as well as a secure means for trusted
partners and vendors to assist with operations and maintenance.

Operation of IT equipment depends on stable electrical power, sufficient ventilation

(or active cooling), as well as a secure location that is safe from unauthorized ac-
cess or exposure to other physical and environmental threats. These dependencies
mean that a highly resilient edge IT installation requires remote monitoring of the in-
frastructure equipment with DCIM software tools. Afterall, you cannot effectively
manage something that cannot be seen. DCIM provides that remote visibility.

Essential Guidance on DCIM for Edge Computing Infrastructure

 Schneider Electric – Data Center Science Center White Paper 281 Rev 0 3

White Paper 280, Practical Guide to Ensuring Availability at Edge Computing Sites,
goes beyond the DCIM system and describes specific actions to take to improve reli-
ability of the power and cooling systems that support the IT.

While traditional DCIM suites offer a wide variety of functions and capabilities, a
smaller subset of core essential functions is needed for local edge environments.
These are summarized in Table 1 below. When you are in the process of select-
ing a DCIM solution, focus on these functions and compare each vendor’s ap-
proach and performance in delivering them. Most, if not all, vendors will offer
these functions in some form. But they will differ in terms of the platform architec-
ture used to deliver the functions. As the next section will show, the platform’s ar-
chitecture drives how effective the DCIM tools will be for local edge environ-
ments.

Table 1
Essential DCIM functions for managing multiple, distributed local edge environments

Function Description Why important

Provides a “read only” connection to all critical Awareness of status changes, trends, and
Device & environmental infrastructure devices (e.g. UPS, rack PDU, alarms, prevents issues from becoming critical
monitoring cooling, etc.) – regardless of vendor – to moni- incidents that could lead to IT service interrup-
tor status, access & alarms in real time. tions.
Provides a means by which infrastructure de- Configuration & updates ensures equipment
Device management vices can be configured and their firmware up- performs as expected and helps secure the
dated. overall system from cyber security threats.
Provides a holistic view of all assets; their loca- IT resiliency requires having an asset inventory
Asset tracking tion, name, status, resource dependencies, etc. and knowing their dependencies.

Raw device data, frequent status change notifi-

Presents useful and actionable information on
Data analytics & cations, and “alarm storms” can overwhelm us-
device status, alarms, and the health of the in-
visualization ers; analytics and clear visualization of data
frastructure systems and their environment.
makes DCIM use simpler and more effective.
Allows DCIM data to be shared with a remote Managed service providers (MSPs) commonly
monitoring and management (RMM) tool or manage edge computing IT and use their own
3 party platform
rd
building management system (BMS) using ap- management platforms; sharing DCIM data with
integration plication programming interfaces (APIs) or an these tools solves “lack of staff” challenge by
SNMP management information base (MIB). enabling trusted partners to manage it for you.

We believe next-generation DCIM platforms are most capable of providing

these functions for local edge applications.

Attributes of A next-generation DCIM platform is defined by 5 key attributes. These items are
what differentiate these modern suites from traditional or legacy DCIM systems that
effective next- were designed for large single-site data centers.
generation DCIM 1. Relies on cloud technologies for ease of implementation, scalability, analyt-
platforms ics, and maintenance
2. Connects to a data lake enabling insight and event prediction with artificial
intelligence (AI)
3. Uses mobile and web technologies and integrates with 3rd party platforms
4. Prioritizes simplicity and intuitive user experiences in its design
5. Serves as a compliance tool to identify and eliminate potential cyber security
risks

Essential Guidance on DCIM for Edge Computing Infrastructure

 Schneider Electric – Data Center Science Center White Paper 281 Rev 0 4

Relies on cloud technologies for ease of implementation, scalabil-

ity, analytics, and maintenance
By hosting the DCIM server in the cloud, deployment is simpler and faster by
eliminating the need to go through the procurement process for a new server for
every site. Next-generation DCIM typically installs as a simple gateway app on an
existing server (physical or virtual). This avoids the often-lengthy security and vali-
dation reviews that can take weeks or months. Since each site would require a
DCIM server, this time savings can be significant when there are dozens or hun-
dreds of small remote sites. This also makes the tool highly scalable in that it can
handle an unlimited number of monitored devices across any number of sites.
Cloud technologies also facilitates further value as described in the attributes below.

Connects to a data lake that enables insights and event predic-

tion with AI
The cloud-based architecture of next-generation DCIM also provides the opportunity
for vendors to offer a “data lake”, a secure repository of massive amounts of anony-
mized device data. “Big data” analytics and machine learning algorithms can
be developed and trained on this data to yield insights and make predictions that
improve reliability, improve efficiency, and/or reduce operating expenses. Early ex-
amples of “big data” analytics and artificial intelligence applied to data center physi-
cal infrastructure include:

• Predicting when UPS batteries will fail – allows for early planning and budget-
ing for service replacements
• Real-time optimization of cooling system controls based on changing climate
and load conditions – reduces operating expenses
• UPS health scorecard sorting the inventory of UPSs based on a determination
of the device’s age and health – simplifies management by first focusing user
on what needs attention most

While this functionality is still in its infancy (at the time of this writing), edge compu-
ting owners and operators considering DCIM solutions today can put them-
selves on the right future path by adopting a next-generation DCIM architec-
ture that includes a data lake.

Uses mobile and web technologies and integrates with 3rd party
platforms
End users, trusted service partners, and vendors can all access the same data at
the same time from any browser or mobile device. Open APIs enables DCIM
data to be shared to MSP RMM tools for easier 3rd party management. Being
browser based and encrypted, the need for VPN and unique login credentials for
every single site is eliminated. This gives real-time visibility to all assets and
sites from one login. These attributes serve to mitigate the challenge of having
many, unmanned sites. For example, mobile access could help remote IT staff
guide untrained, on-site personnel to troubleshoot and resolve issues without dis-
patching service. And integration into your MSPs RMM tool means that they can
now manage and service your physical infrastructure equipment for you, just
as they might be doing for your IT applications.

Essential Guidance on DCIM for Edge Computing Infrastructure

 Schneider Electric – Data Center Science Center White Paper 281 Rev 0 5

Prioritizes simplicity and intuitive user experiences in its design

Next-generation DCIM tools tend to perform better than legacy DCIM suites in terms
of ease of installing, configuring, and using the software. Some of the common im-
provements include things like:

• Installations use easy-to-follow wizard-based routines

• Device alarm thresholds come with useful default settings
• Device health scorecards sort devices in need of attention or action first
• Performance benchmarking provides context on how you are performing rela-
tive to peers
• Alarms and status changes are grouped based on common causes to elimi-
nate alarm “storms”
• Device setting and policy changes can be mass applied to many devices at
once
• DCIM app gateway and device firmware can be set to auto-update to roll out
bug fixes, feature enhancements, and security patches as soon as they are
available; no longer a vendor-provided server that must be maintained by the
end user

Serves as a compliance tool to identify and eliminate potential

cyber security risks
Given that DCIM systems are made up of software apps, servers, gateways, and
critical infrastructure devices, all inter-connected over mobile and IT networks, it is
important to ensure cyber security best practices are continuously followed by both
the vendor and end user. Next-generation DCIM should simplify this for the end
user by automating the detection and reporting of DCIM gateway and device
vulnerabilities. Some DCIM solutions do this using a threat assessment tool. Us-
ers are notified if device configurations (e.g., set to use SSH or Telnet) put the de-
vice at risk of attack. Devices with outdated firmware is also identified. This greatly
simplifies management and automates a critical function of the DCIM system.

Tips & guidance Like any enterprise grade software suite, successful DCIM implementations require
organizational buy-in and on-going cooperation and participation amongst key stake-
for getting started holders. While DCIM ultimately aims to simplify and, to some degree, automate
with DCIM at the management of data center infrastructure, the users of the system must do their part
to ensure the value of the software is realized. For example, the operations and
edge maintenance (O&M) of the software system must be built in to the organization’s
O&M program. The facility O&M program’s change management processes must be
adapted to account for the DCIM system. This takes commitment and continuous
effort by management and operations teams. If this is not done, the implementation
and use of DCIM can fail. Note that next-generation DCIM O&M requirements are
less burdensome than legacy DCIM. Some of the key DCIM-related processes that
need to be accounted for in the existing IT or facility O&M program include:

• configuring device network settings for equipment that is added or replaced,

and confirming that network communication is established
• reviewing status of device firmware periodically and updating when available
• ensuring device alarm thresholds and notification policies are set properly
• reviewing DCIM alarms and status changes regularly

Essential Guidance on DCIM for Edge Computing Infrastructure

 Schneider Electric – Data Center Science Center White Paper 281 Rev 0 6

• maintaining DCIM software, server and/or gateway with updates, bug fixes,
and security patches (NOTE: most next-generation DCIM solutions can auto-
mate this process and there’s no server to maintain)

White paper 196, Essential Elements of Data Center Facility Operations, describes
the importance of change management process and DCIM in operating a data cen-
ter efficiently and effectively. White paper 170, Avoiding Common Pitfalls of Evalu-
ating and Implementing DCIM Solutions, goes into much more detail on the subject
of integrating DCIM processes into your existing operations program.

Table 2 provides a summary list of tips to help make your deployment and opera-
tions of DCIM a success.

Table 2
Advice on how to better ensure the value of DCIM at edge compute sites is realized
Tips for ensuring DCIM
Description
success at the edge
Obtain buy-in and commitment Without the full support of everyone involved, the system can fall out of use. Be
from key stakeholders before sure to include the operations team and/or the MSP to assist in selecting and im-
purchase and implementation plementing the tools. (See next section for more information about role of MSP)

Cloud-based DCIM platforms are better architected for managing highly distrib-
Select a next-generation solution uted, unmanned infrastructure, and puts you on the right path to more predictive,
automated management.

The most critical function is being aware of alarms and status changes with the
physical infrastructure. Once this is in place and the processes to use and main-
Focus on monitoring first
tain it are mature, then other functions and features of DCIM can be considered.
(See next sub-section below)

Integrate processes to operate Operating and maintaining a DCIM solution takes effort and discipline. It must be
and maintain DCIM with your ex- an official, documented part of the existing O&M program to minimize the risk that
isting O&M program(s) the tools fall out of use.

Avoid the “information problem” by using next-generation DCIM

Regardless of who is doing the day to day management, it is important to choose
software management tools that clarify and prioritize status notifications and alarms.
Traditional tools might leave operators spending a lot of time trying to evaluate each
one. What does it mean? Do I need to do something? Instead of providing raw
data about the IT environment and supporting infrastructure, modern tools draw con-
clusions to either provide a root cause for the problem or to make clear which
alarms or devices are most critical and require attention. For example, a retail store
chain may have thousands of UPS units deployed across the country. The manager
could see hundreds of thousands of data points and status change notices from that
population. It would be easy for critical alarms to go unnoticed or for a storm of
alarms to occur with no understanding of what is driving them. Effective software
tools focus the operations manager on only the UPS units that need attention first or
on the one alarm that started a cascade of further alarms. One way to do this is
through a device health score card system that ranks devices based on their health.
For a UPS, the health would be determined by the age of the unit, battery charge
capacity, temperature, and so on. Figure 2 shows an example “health” score card
for a fleet of UPSs.

Essential Guidance on DCIM for Edge Computing Infrastructure

 Schneider Electric – Data Center Science Center White Paper 281 Rev 0 7

This capability saves time and reduces trial and error efforts to understand what is
happening with the equipment. It reduces the need to send out service personnel to
investigate problems. It reduces cases where equipment thought to be faulty is
prematurely replaced with new. We often hear of cases where service people are
dispatched to replace a “bad” UPS only to find it was being overloaded by store em-
ployees who found the unused outlets useful for powering space heaters, vacuums,
or coffee makers. Effective software tools would’ve clearly alerted management that
the UPS had shut down because they had been overloaded, not because they had
failed.

Figure 2
An example screenshot
showing a health scorecard
for a fleet of managed
UPSs; units needing the
most attention are sorted to
the top of the list.

Confidential Property of Schneider Electric | Page 33

To monitor yourself or rely on partners and/or vendors

DCIM was originally designed to be self-managed and monitored by the owner of
the tools. You buy a lifetime license and operate and maintain the system yourself.
Vendors typically offered implementation and startup services as well as training
programs to ensure your operations team could handle it for the long term. Remote
monitoring services by the DCIM vendor or other 3rd party service providers have ex-
isted, but they tend to be very basic and reactive in nature. Essentially the infra-
structure devices are configured to send alarms to the vendor who react if some-
thing goes wrong.

Next-generation DCIM provides a platform that makes it easier for 3rd parties to
manage and monitor for you or together with you. The end user, the MSP, and the
DCIM vendor can all have access to the same data at the same time. Under this
new digital services model, servicing and replacing components could be carried out
even before the local site knows there’s an issue. Enabling full use of the DCIM
system by 3rd parties mitigates the key edge computing challenge of having un-
manned sites. Your service providers and DCIM vendor can, in effect, become your
virtual staff. Connecting your MSP to the system can happen either by giving them
access to the DCIM software itself or by using application programming interfaces
(APIs) so that MSPs can bring DCIM information in to their RMM software tools.

However, as explained in previous sections, Modern DCIM tools are making it easier
to use the software yourself if that is the preferred operating model. Using a simple
gateway app instead of having to source and implement a 3rd party server simplifies
getting started. Simple wizard-based setup and configuration tools with suggested
default settings makes implementation easier. And a move towards providing more

Essential Guidance on DCIM for Edge Computing Infrastructure

 Schneider Electric – Data Center Science Center White Paper 281 Rev 0 8

intuitive, insightful analytics vs. raw data also contributes to making next-generation
DCIM tools simpler to do yourself. Mobile apps also help simplify management by
giving managers and operators visibility from anywhere, at any time.

Ultimately, you have 3 basic choices in terms of who manages the DCIM system:

• Do it all yourself
• Let the DCIM vendor remotely monitor in collaboration with you
• Leverage an MSP who is managing the IT (Note, the MSP might also benefit
from having the DCIM vendor remotely monitor as well in this case.)

The service of having the DCIM vendor remotely monitor the infrastructure devices
is commonly called digital remote monitoring. The benefit of using this type of ser-
vice is two-fold:

• Expertise –The vendor’s operators are experts on the equipment being moni-
tored
• Time – Time to detect and troubleshoot problems is significantly reduced, and
the need for the end user to have to call vendor’s tech support and navigate to
the right person is eliminated; Time to recover is faster in that the vendor can
immediately dispatch field service

To determine the right choice, you should ask yourself the following questions:

• How critical is the IT infrastructure?

• Do you already have or plan to use an MSP to manage the IT?
• Does your operations team have the expertise and bandwidth to monitor and
manage the physical infrastructure?
• Do you require less than 24hr resolution of problems and replacement of failed
devices?
• What service contracts exist for the infrastructure hardware already?

Essential Guidance on DCIM for Edge Computing Infrastructure

 Schneider Electric – Data Center Science Center White Paper 281 Rev 0 9

Conclusion Power, cooling, and environmental/security monitoring equipment is critical to the

continued operation of the IT at edge computing sites. Particularly since these are
often geographically dispersed and unmanned, next-generation DCIM software that
relies on cloud, mobile, and AI technologies should be used to effectively monitor
this infrastructure from afar. Compared to legacy DCIM solutions, modern DCIM
tools are easier to scale, use, maintain, and provide a way for MSPs to remotely
manage for you. Newer DCIM platforms are also beginning to take advantage of AI
technologies that will make management more predictive and automated.

We believe well-maintained and operated DCIM systems make data centers more
reliable and efficient. This means having buy-in from all stakeholders, integrating
DCIM processes with the existing O&M program, and either having the discipline to
regularly monitor the system yourself or choosing to have your trusted partner or
vendor manage it all for you.

About the author

Patrick Donovan is a Senior Research Analyst for the Data Center Science Center at Schneider
Electric. He has over 25 years of experience developing and supporting critical power and cooling
systems for Schneider Electric’s IT Business unit including several award-winning power protec-
tion, efficiency and availability solutions. An author of numerous white papers, industry articles,
and technology assessments, Patrick's research on data center physical infrastructure technolo-
gies and markets offers guidance and advice on best practices for planning, designing, and opera-
tion of data center facilities.

RATE THIS PAPER 

Essential Guidance on DCIM for Edge Computing Infrastructure
 Schneider Electric – Data Center Science Center White Paper 281 Rev 0 10

Classification of Data Center Infrastructure Management (DCIM) Tools

Resources White Paper 104

Solving Edge Computing Infrastructure Challenges

White Paper 277

Practical Guide to Ensuring Availability at Edge Computing Sites

White Paper 280

Essential Elements of Data Center Facility Operations

White Paper 196

Avoiding Common Pitfalls of Evaluating and Implementing DCIM Solutions

White Paper 170

Browse all
white papers
whitepapers.apc.com

Browse all
TradeOff Tools™
tools.apc.com

Contact us
© 2019 Schneider Electric. All rights reserved.

For feedback and comments about the content of this white paper:
Data Center Science Center
[email protected]

If you are a customer and have questions specific to your data center project:
Contact your Schneider Electric representative at
www.apc.com/support/contact/index.cfm

Essential Guidance on DCIM for Edge Computing Infrastructure