Improve data center

operations with
forecasting
Leverage asset data to get the most out of your DCIM platform.
siemens.com/dcim
Improve data center operations with forecasting

Introduction

There’s a lot that goes into managing a data center. Rising complexity
due to the thousands of applications and devices plus the absence
of knowledge-transfer between stakeholders makes daily management
an operational challenge. Add planning for future deployments
and tenants to the mix and the challenge can easily become a costly,
haphazard nightmare. Data centers are critical components of modern
enterprises and should be used to drive business, not hinder it.
In these environments, it is not only essential to have continuous
monitoring in place, but also to effectively manage capacity and
discover hidden potential.

The usual answer to these challenges is the implementation of a

data center infrastructure management (DCIM) platform. While
a typical DCIM solution can bring valuable insights, most don’t go far
enough. DCIM software generally focuses on providing real-time
monitoring and a consolidated view of data with current inventory.
We cannot take these functionalities for granted, but the true challenge
arises when insights into future states are required. Knowing that in
this moment the data center is using 80 % of installed capacity is
valuable, but what about a month from now? Or a year? How does
knowing current capacity help when we need to know if we can
support 10 new tenants who each need 100 kW in six months time,
while minimizing overprovisioning?

The answer is forecasting. Today’s DCIM solutions need to do more

than asset tracking – they need to support business objectives and
enable all stakeholders in a data center to make risk-free decisions,
whether their focus is facilities, IT, or the overall health of the enterprise.

The following discussion will present why a true DCIM requires not
only capacity management functionalities but also must accurately
assess current and future states within a data center.

2
3
Improve data center operations with forecasting

Challenges: Why forecasting

is a game-changer
The data center industry has come to a point where traditional inventory manage-
ment and monitoring do not address the growing need for scalability and actionable
information. Increased demand is putting more pressure on IT and facilities alike.
In one report, it was noted that almost half of data center decision makers need to
increase power and cooling capacity to guarantee service delivery. Among those
decision-makers, 24 % reported a loss of clients directly related to the lack of quality
operational data [1]. In a market that’s projected to grow by $284 billion by 2023 [2],
complete visibility into data center operations is critical for business continuity,
and represents a large opportunity to eliminate waste, increase utilization, and save
money. Two key challenges stand in the way of driving business objectives while
maintaining service continuity:

1. Overprovisioning

Lack of insight into existing and future installations has made overprovisioning the
only option available for data center managers who need to maintain service conti-
nuity. It not only represents a heavy capital investment but also higher administra-
tion, staffing, and maintenance costs. In one assessment, it was found that due
to overprovisioning 136 % of operation budget was required to run the average data
center, and that 3 data centers are built for every 2 that we need [1].

It’s clear that this issue is pervasive, and that a simple asset inventory will not miti-
gate its effects. Imagine instead being able to predict when capacity will be
released, measure the impact of new tenants and projects on existing infrastruc-
ture, and therefore plan effectively for incoming clients and installations. Instead
of having reactionary, incident-driven strategies, a proactive approach can avoid
unnecessary CAPEX investments and extend the lifecycle of existing assets. Action-
able information needs to be available to all stakeholders that enables them to
make the best decisions, rapidly and confidently.

2. Tenant management

For this discussion, we define tenant management as the management of infra-

structure with consideration of the end-user to whom the resources are assigned.
The end-user can be an internal group or a colocation tenant.

Ensuring that all end-users have continuous service and budgeting resources for
current and incoming requests is a costly, time-con­suming process. There are two
primary issues when it comes to tenant management:

4
a) Firstly, when a request for resources comes along, it’s important to know if
there’s available capacity. Taking on a new client or project only to find a
capacity shortfall is not only expensive but also stressful.

b) Secondly, not knowing when the biggest contracts are going to expire and their
effect on the data center space blocks the use of valuable capacity. It’s important
to plan how to use and, more importantly, sell space that will become available.

In short, effective planning of resources requires insight into the full lifecycle of
datacenter assets, services, and users.

The best way to counter the above challenges is with a DCIM with integrated fore-
casting. DCIM alone tends to focus primarily on asset management and doesn’t
permit effective planning. On the other hand, even the best forecasting tool needs
to be driven not only by contract and project timelines but also by real-time data
and asset management information.

Thus, a holistic DCIM solution is one that includes forecasting capabilities, and
therefore permits the management of upcoming deployments as well as current
installations. This enables the development of insightful capacity metrics that
support business objectives, rendering the data center profitable.

5
Improve data center operations with forecasting

Forecasting: Asset
­management amplified
To have optimal business value, a comprehensive asset management solution
needs to be structured, provide visibility, and deliver reliable data to decision-­
makers. Most importantly, it should provide a full lifecycle view of all resources
and show how the data center evolves over time.

A full lifecycle view not only tracks resources over their lifetime, but also details
the applications and tenants supported by those assets. To do all of this simultane-
ously, it’s necessary to go beyond traditional asset management and enhance
the analysis with added tiers of information, such as the date of decommissioning,
contract start and expiry date, associated end user, and more. Once this data is
consolidated, powerful forecasting reports that show the evolution of the data
center can be constructed. These analyses form the foundation of forecasting.

So, what is forecasting exactly? For the scope of this discussion, forecasting is
defined as follows:

The use of past, current, and future operational data to monitor and predict the
impact of data center deployments on capacity metrics, including but not limited
to power, space, and cooling usage. A true forecasting solution gives insight into
current usage and allows operators to construct predictive analyses based on
“extended” metrics, such as revenue and cost per rack or tenant. A true forecasting
solution enables data center operators to answer the following:

• What is my current capacity and when will I run out of resources?

• How many new projects can I accommodate today? When will I be able to
accommodate more?
• Do I have enough capacity to accept new tenants and future expansions of
existing tenants and/or applications?

In addition, to be advantageous and profitable, the results rendered from the

analysis must enable rapid, risk-free decision-making. The analysis cannot stand
alone and needs to be intuitive and easy to understand. To effectively leverage
forecasting, there are a few essential components that greatly aid in data
processing:

• Ability to see, at any point in time, if capacity is being used to its optimal
• potential.
• Ability to automatically generate reports
• of aggregated data showing current and future capacity data.
• Ability to visualize data in a meaningful way.
• The above capabilities can be leveraged
• in a variety of use-cases. This discussion will focus on two: Capacity Planning
and Supporting the Sales Cycle.

6
7
Improve data center operations with forecasting

Operational use-cases
Capacity planning
The challenge:

As we’ve seen, traditional capacity planning gives an overview of existing assets.

For example, in a high-performance computing (HPC) Infrastructure, a monitoring
and inventory system might tell us how many racks are installed, how full they are,
and how much power they’re consuming. With this information, we understand
how much capacity we have remaining in real-time. When it comes to expansions,
however, this information does not help us understand the impact of future
deployments on capacity, or when expansions will be necessary. Here, the chal-
lenge is that there’s a general picture of the current state of affairs in the data
center, but it’s difficult to correlate the data center’s current state with the capacity
requirements of incoming projects, and therefore complicated to plan for
upcoming deployments. This represents a large risk. There’s no way to know if
there is an urgent need to pour money into procuring more capacity, or if current,
occupied resources will be released and therefore be available for future applica-
tions. Simply put, there’s no visibility into the impact of upcoming installations on
KPIs and critical equipment.

The solution:

A tool to solve the challenges above requires several components. First, operators
need to shift their focus from just current capacity to current capacity plus poten-
tial capacity (i.e. not just “There are 400 racks installed,” but “There are 400 racks
installed with a potential of 550”). This simple change already allows data center
operators to expedite decision-making and provisioning – they know exactly where
there is room to grow and can assign capacity when needed.

8
9
Improve data center operations with forecasting

Next, a complete forecasting tool should allow operators to reserve capacity for
­different applications, over specific periods. ­Allocated capacity can then be aggre-
gated by the tool for any given point in time. Then, by associating the timeline
of proposed projects, expansions, and tenants with ­specific resources or quantity
of capacity, operators know:

1. The amount of resources allocated to particular applications

2. When that capacity is going to be freed
3. When more capacity will be required

A complete forecasting tool should be able to perform this type of analysis and
tell operators when capacity that is currently being used will become available
again, allowing them to effectively allocate existing capacity. With this data, a
timeline can be constructed showing how the data center will evolve over time,
giving maximum visibility into operations for years to come.

Finally, a true forecasting tool also allows operators to consider asset lifetime.
Overlaying the amount of capacity used, contract (or project) expiration date,
and the predicted asset decommissioning date gives a powerful overview of how
the data center will grow and change over time.

Benefits:

1. Anticipate the utilization of key capacity metrics

2. Use actionable data to avoid delays in approval and execution of expansion plans
3. Optimize utilization of existing installations

Figure 1: By forecasting capacity usage, operators can predict capacity shortfalls far into the future. The image above shows
current power allocated, reserved power capacity, and forecasted power needs over a span of 9 years.

10
11
Improve data center operations with forecasting

Operational use-cases
Supporting the sales cycle
The challenge:

Current capacity planning and inventory tools don’t give a full picture of operations
over time, preventing effective allocation of resources. Without this information,
sales managers do not have a panoramic view of when contracts expire (in other
words, capacity over time), and therefore can’t identify sellable space, leading to
costly overprovisioning for upcoming contracts. Planning and critical decisions are
consequently slow, costly, and based on unreliable data.

Solution:

Leveraging a forecasting solution can help sales managers gain insight into opera-
tional data and the data center’s evolution over time. A comprehensive forecasting
tool gives the ability to correlate the timeline of a project or the duration of a
contract to a distinct set of resources. When that timeline finishes, that set of
resources becomes available.

Figure 2: Traditional Capacity Planning gives an over-

Today view of current capacity usage and how much capacity
Sellable capacity Allocated capacity is assigned to each end-user.

Figure 3: Leveraging a forecasting-enabled DCIM

solution can help identify sellable space long before it
becomes available, allowing for the efficient allocation
of resources and preventing costly overprovisioning. In
the image above, allocated capacity refers to capacity
that is already assigned to and will be in use by a tenant,
whereas reserved capacity might be occupied today but
has already been put aside for another tenant in the
Six months from now future. Sellable capacity is capacity that will be available
Sellable capacity Allocated capacity Reserved capacity after some time but may be currently occupied.

12
 While that sounds trivial, structuring this information in an intuitive way delivers
incredible value, as managers can see how and where to allocate resources far into
the future and can identify sellable space even if it’s currently occupied. In addi-
tion, it’s possible to predict when capacity will finish, eliminating costly over-provi-
sioning for upcoming tenants. With minimal input, a powerful and advantageous
outcome can be achieved.

Furthermore, extracting the business-critical data associated with contracts, such

as revenue, enables a powerful analysis in which the value of every resource in the
data center can be assessed. It is then possible to map projected profits as they
relate to data center resources, putting the business intelligence required to make
provisioning and resource-allocation in the hands of decision-makers.

Benefits:

1. Identify available, sellable space

2. Visualize and be prepared for when your biggest contracts are going to expire
3. Uncover and optimize capacity allocation for existing tenants

Figure 4: Linking resources to specific uses or contracts enables powerful analyses of the overall health of the data center. In the image
above, secured revenue refers to revenue to-be from finalized contracts, whereas the planned revenue is the result of a forecasting
analysis and refers to revenue that could be incurred by selling space that will become available. We can see that the overall health of the
business is not great in the case above!
Improve data center operations with forecasting

The future of DCIM

and forecasting
Forecasting-enabled DCIM offers many ­opportunities to reduce costs and increase
profitability. By gaining visibility into capacity usage over time, data center opera-
tors can plan effectively and maximize utilization. In turn, they can reduce over­
provisioning and anticipate the effect of upcoming deployments on critical infra-
structure. For sales managers, forecasting means knowing and preparing for when
contracts are going to expire and the ability to identify and sell available space.
When combined with ­network power and cooling infrastructure, today’s newer
DCIM software systems provide the necessary tools for efficient and effective
management “out of the box”.

Complexity is now manageable. When it comes to capacity planning, visibility in

real-time is not enough to combat the uncertainty and lack of information that
prevent effective decision-making. In today’s data center environment, planning
is crucial. Luckily, effective solutions such as forecasting are already available
to transition us to the next phase of data center management, where insight into
the future is critical for business continuity.

References

[1]
Datacenter Dynamics, “The True Cost of Data Center Over-Provisioning,”
Datacenter Dynamics, 2019.
[2]
Business Wire, “Global Data Center Market Outlook 2019-2023,” 2019.
[3]
eWeek, “Report Finds Over-Provisioned Hardware an Expensive
IT Problem,” 2018.

14
Published by Smart Infrastructure intelligently connects energy
Siemens Switzerland Ltd systems, buildings and industries to adapt and evolve
Smart Infrastructure the way we live and work.
Global Headquarters
Theilerstrasse 1a We work together with customers and partners
6300 Zug to create an ecosystem that intuitively responds
Switzerland to the needs of people and helps customers to better
Tel +41 58 724 24 24 use resources.

For the U.S. published by It helps our customers to thrive, communities to

Siemens Industry Inc. progress and supports sustainable development.
100 Technology Drive
Alpharetta, GA 30005 Creating environments that care.
United States siemens.com/smart-infrastructure

Subject to changes and errors. The information given in this document only contains general descrip-
tions and/or performance features which may not always specifically reflect those described, or
which may undergo modification in the course of further development of the products. The requested
performance features are binding only when they are expressly agreed upon in the concluded contract.
© Siemens 2021