Data Center:

Standardization & Modularity

School of Computer Engineering

KIIT Deemed to be University
Bhubaneswar
UNIT-2: Data Center: Standardization and
Modularity
▪ Data Center (DC)

▪ DC Configuration

▪ DC Standards

▪ Benefits of Standardization

▪ Power of Modularity

▪ Challenges

Unit-2 : Cloud Computing School of Computer Engineering 2

 #1: Data Center
▪ A data center is a physical location that stores computing machines and their related
hardware equipment. It contains the computing infrastructure that IT systems require,
such as servers, data storage drives, and network equipment. It is the physical facility
that stores any company's digital data. [Amazon]

▪ A data center is a physical room, building or facility that houses IT infrastructure for
building, running, and delivering applications and services, and for storing and managing
the data associated with those applications and services [IBM]

▪ A facility used to house computer systems and

associated components. [Wikipedia]

▪ Collection of physical compute, storage, and network

resources

Unit-2 : DOS School of Computer Engineering 3

 #2 Data Center Configuration
▪ Most enterprise data center infrastructure falls into three broad
categories:

▪ Compute
▪ Storage
▪ Network

▪ Also, data center equipment includes support infrastructure like

power systems, which help the main equipment function effectively.

Reference: https://aws.amazon.com/what-is/data-center/
Unit-2 : Cloud Computing School of Computer Engineering 4
 #2 Data Center Configuration (contd…)
Computing infrastructure
▪ Computing resources include several types of servers with varying internal memory,
processing power, and other specifications. We give some examples below.

1. Rack servers
▪ Rack servers have a flat, rectangular design, and you can stack them in racks or
shelves in a server cabinet. The cabinet has special features like mesh doors, sliding
shelves, and space for other data center resources like cables and fans.

2. Blade servers
▪ A blade server is a modular device and you can stack multiple servers in a smaller
area. The server itself is physically thin and typically only has memory, CPUs,
integrated network controllers, and some built-in storage drives. You can slide multiple
servers into a storage unit called a chassis. The chassis facilitates any additional
components that the servers inside it require. Blade servers take up less space than
rack servers and offer higher processing speed, minimal wiring, and lower power
consumption.
Reference: https://aws.amazon.com/what-is/data-center/
Unit-2 : Cloud Computing School of Computer Engineering 5
 #2 Data Center Configuration (contd…)
Storage infrastructure
▪ The following are two types of data center storage systems.

1. Block storage devices

▪ Block storage devices like hard drives and solid-state drives store data in
blocks and provide many terabytes of data capacity. Storage area
networks (SANs) are storage units that contain several internal drives and
act as large block storage systems.

2. File storage devices

▪ File storage devices, like network-attached storage (NAS), can store a
large volume of files. You can use them to create image and video
archives.

Reference: https://aws.amazon.com/what-is/data-center/
Unit-2 : Cloud Computing School of Computer Engineering 6
 #2 Data Center Configuration (contd…)

Network infrastructure
▪ A large number of networking devices, such as cables, switches,
routers, and firewalls connect other data center components to each
other and to end-user locations. They provide flawless data
movement and connectivity across the system.

Reference: https://aws.amazon.com/what-is/data-center/
Unit-2 : Cloud Computing School of Computer Engineering 7
 #2 Data Center Configuration (contd…)
Support infrastructure
Data centers also contain these components:
▪ Power subsystems
▪ Uninterruptible power supplies (UPS)
▪ Backup generators
▪ Ventilation and cooling equipment
▪ Fire suppression systems
▪ Building security systems
▪ These data center components support the main equipment so that
you can use the data center facilities without interruption.

Reference: https://aws.amazon.com/what-is/data-center/
Unit-2 : Cloud Computing School of Computer Engineering 8
#3 Standards in Data Center Design
▪ Data centers increased in size and complexity and began to store
sensitive and critical information, governments and other
organizations-imposed regulations on them

▪ Telecommunications Industry Association (TIA) established four

levels or standards that cover all aspects of data center design,
including
• Architecture and topology
• Environmental design
• Power and cooling systems and distribution
• Cabling systems, pathways, and redundancy
• Safety and physical security

Reference: aws.amazon.com
Unit-2 : Cloud Computing School of Computer Engineering 9
#3 Standards in Data Center Design(contd…)
Tier I
▪ A Tier I data center is the basic capacity level to support IT systems
for an office setting and beyond. Some of the requirements for a Tier
I facility include:
• Uninterruptible power supply (UPS) for power outages and spikes
• A physical area for IT systems
• Dedicated cooling equipment that runs 24/7
• A backup power generator

▪ Tier I protects against service disruptions from human error but not
against unexpected failure or outage.
▪ Annual downtime of 29 hours in Tier I data centers.
Reference: aws.amazon.com
Unit-2 : Cloud Computing School of Computer Engineering 10
#3 Standards in Data Center Design(contd…)
Tier II
▪ Tier II facilities provide additional cooling components for better
maintenance and safety against disruptions.
▪ Tier II data centers must have the following:
• Engine generators
• Chillers
• Cooling units
• Pumps

▪ Although you can remove components from Tier II data centers without
shutting them down, unexpected failures can affect the system.
▪ Annual downtime of 22 hours from a Tier II data center.
Reference: aws.amazon.com
Unit-2 : Cloud Computing School of Computer Engineering 11
#3 Standards in Data Center Design(contd…)
Tier III
▪ Tier III data centers provide greater data redundancy, and you can
maintain or replace equipment without system shutdown.
▪ They also implement redundancy on support systems like power and
cooling units to guarantee only 1.6 hours of annual downtime
Tier IV
▪ Tier IV data centers contain several physically isolated systems to
avoid disruption from both planned and unplanned events.
▪ They are completely fault-tolerant with fully redundant systems and
can guarantee a downtime of only 26 minutes each year.
Reference: aws.amazon.com
Unit-2 : Cloud Computing School of Computer Engineering 12
 #4 Benefits of Standardization
▪ Standardization refers to the consistent use of predefined components, practices, and
layouts across the facility.
▪ Standardization allows for several foundational benefits that justify uniformity in modern
data center design:
• Affordability: Standardized components (racks, servers, and layouts) lead to
economies of scale, meaning cheaper procurement, easier deployment, and
streamlined maintenance. You buy in bulk, deploy faster, and fix things quicker.
• Efficiency: Standardized practices make operations predictable and optimized.
Everyone knows the drill, resulting in less wasted time and resources. If and when
problems arise, the approach to resolving them is recognized and repeatable.
• Scalability: Standardized building blocks enable easy expansion and adaptation to
changing needs. The data center grows smoothly without major headaches and is
easily configured at its new scale.
• Reliability: When everything follows the same blueprint, errors are minimized.
Consistent practices and components translate to fewer hiccups and more uptime.
Think of it as building your data center with strong, reliable bricks

Reference: https://www.linkedin.com/pulse/why-standardization-modularity-essential-data-center-design-8zwzc/
Unit-2 : Cloud Computing School of Computer Engineering 13
 #5 Power of Modularity
▪ Imagine your data center built like a set of interconnected blocks instead of a monolithic
structure. The concept of modularity revolves around the construction and operation of a data
center using self-contained, interchangeable units. The benefits are substantial:

• Flexibility: With modular systems, simply add prefabricated units like building blocks for rapid
deployment and adaptation.

• Serviceability: Modular design makes replacing or upgrading components a breeze. Just swap
one component for another without impacting the entire system. It’s ideal for creating
redundancies and failsafe protocols.

• Resilience: System failures will happen, but with modularity, they’re isolated. A faulty “block”
can be replaced quickly while the rest of the data center keeps humming. Downtime becomes
less of a worry.

• Future-proof: Technology evolves rapidly, but modularity adapts. New modules can be
integrated seamlessly, allowing a data center to implement emerging technologies
Reference: https://www.linkedin.com/pulse/why-standardization-modularity-essential-data-center-design-8zwzc/
Unit-2 : Cloud Computing School of Computer Engineering 14
 #5 Power of Modularity (contd…)
▪ Reduced Human Error: All operations from assembly of the modules to the troubleshooting of
the system, documentation and training are easier and more effective, ensuring the proficiency
of the employees and preventing errors.

▪ Prediction of Problems: The working principles and the standardized programs for modular
data centers such as equipment monitoring and predictive maintenance programs form a
powerful defense mechanism against “unexpected” incidents.

▪ Improved Efficiency: Efficiency is improved as the learning effect spreads and is promoted
among employees. More knowledgeable employees are better at solving problems. Reducing
human error also reduces the time needed for problem solving including answering
troubleshooting calls on hot lines. This releases human resources, which can be used more
efficiently.

▪ Advantages of Batch Production: Standardization of parts and processes makes the batch
production of modular data centers possible. Batch production has the advantages of lower
costs, higher quality, easier repair and shorter lead time.
Reference:
https://www.deltapowersolutions.com/en-in/mcis/technical-article-modular-data-centers-system-architecture-and-benefits.php
Unit-2 : Cloud Computing School of Computer Engineering 15
 #5 Power of Modularity (contd…)
▪ Expandability of modular data centers: A modular data center can be
deployed depending on current IT requirements and can be expanded in the
future, if needed. TCO is reduced significantly as a result.

▪ Flexibility of the modular system: When a modular data center is installed,

upgraded and reconfigured, the independent components, standard interfaces
and comprehensive structures save both time and cost.

▪ Reduction of failure repair time: The flexibility and plug-in characteristics of

the modular data center allows many operations to be completed in the factory
before delivery, such as pre-cabling of the distribution equipment, as well as
after delivery, such as the repair of a power module. Statistically, the failure rate
of operations carried out in the factory is much less than operations carried out
at the site. Compared with a UPS power module repaired at the site, the
probability that a module repaired in the factory brings about a power or other
failure or cannot recover its full-load capability is hundreds of times less.
Reference:
https://www.deltapowersolutions.com/en-in/mcis/technical-article-modular-data-centers-system-architecture-and-benefits.php
Unit-2 : Cloud Computing School of Computer Engineering 16
 #6 Challenges
▪ The data center of the future doesn’t just employ standardization and modularity. The key lies in
harmonizing these concepts. While both offer significant benefits, overdoing either can come
with drawbacks.

▪ Strict standardization can hinder innovation and adaptability. Likewise, standardizing everything
may stifle exploring new technologies or processes. Unconstrained modularity also creates
pitfalls. Too much modularity can lead to increased complexity in managing and maintaining
diverse components. Meanwhile, highly customized modular systems might be more expensive
due to lower economies of scale compared to standardized components.

▪ Striking a balance between standardization and modularity comes down to understanding their
limitations while leveraging their benefits. For example, data center designers can strive to
combine the efficiency and cost-effectiveness of standardization with the flexibility and future-
proofing of modularity.

▪ Data centers are complex systems, and the optimal balance will vary depending on the
organization’s specific needs and goals. Finding the sweet spot between standardization and
modularity is crucial for building efficient, scalable, and resilient data centers ready for the future
Reference: https://www.linkedin.com/pulse/why-standardization-modularity-essential-data-center-design-8zwzc/
Unit-2 : Cloud Computing School of Computer Engineering 17
 Thank You!

Unit-2 : Cloud Computing School of Computer Engineering 18