Datacenter, Networking
and Terminologies
Week # 03
What is Datacenter
What is Datacenter networking
Datacenter topology
Agenda Applications
Scalability, security, reliability,
Issues in Datacenter cooling system, power consumption
Software defined Datacenter?
Examples Facebook, Google, Microsoft
Data Center
• A data center -- also known as a datacenter or data centre -- is a facility
composed of networked computers, storage systems and computing
infrastructure that organizations use to assemble, process, store and
disseminate large amounts of data.
• A data center or data centre is a building, a dedicated space within a
building, or a group of buildings used to house computer systems and
associated components, such as telecommunications and storage systems.
• A data center is the physical facility providing the compute power to run
applications, the storage capabilities to store data, and the networking to
connect employees with the resources needed to do their jobs.
Data + Center
Huge collection of networking devices + computing
devices + storage devices
Datacenter Networking devices Switches, routers
Computing devices Servers
DAS
Storage devices NAS
SAN
RACKS
Servers
Routers
Datacenter
topology Switches
Cables
Virtual Machines
Data Center Topologies
Data Center Topologies
• Rack mountable Servers
• Blade server
• Switches are available in different
throughputs or speeds, the rate they
transmit data in megabits per second
(Mbps). For example, fixed-configuration
switches can provide Fast Ethernet
(10/100 Mbps), Gigabit Ethernet
Switch (10/100/1000 Mbps), Ten Gigabit
(10/100/1000/10000 Mbps), and even
speeds 40/100 Gbps (gigabits per second)
speeds. The switch speed you choose
depends on the type of throughput you
need. If, for example, you need to move
large data files on a regular basis, you
should consider a Gigabit Ethernet
switch.
Number of
ports
As is the case with switch
speed, the number of ports
available in a switch can vary.
The bigger your small business
and the more network users you
have, the more ports you'll
need. Fixed-configuration
switches are usually available
with five, eight, 10, 16, 24, 28,
48, or 52 ports.
Power over Ethernet (PoE) allows you to power a
device like an IP phone or wireless access point
over the same cable as your data traffic. If you have
PoE vs. non- a larger network, PoE can provide you with great
flexibility by allowing you to place endpoints
PoE anywhere in the office. This is especially handy in
spaces where it’s difficult to run a power outlet.
Small, 4 racks per unit
Types Medium, 10 racks
Large, 20 racks
Power consumption
Small: < 1
Rapid
MW, 4 racks
deployment
per unit
Medium:
Built-in 1-4 MW,
cooling, 10 racks
per unit
Large: > 4
MW, 20
racks per
unit
Container
based
Ready to Use.
Built in cooling. Easy to scale.
Data Center trailer parks.
Suitable for disaster recovery, e.g., flood,
earthquake
Offered by Cisco, IBM, SGI,
Sun/ORACLE,…
Ref: Datacenter Infrastructure – mobile Data Center from Emerson Network Power
, http://en.m-info.ua/180-container-data-center/755-datacenter-infrastructure-mobile-data-center-from-emerson-network-power
Central to the Business
business critical apps
Design Proper
Plug n play
Principles planning
Proven
layered
approach
Design Principles
The networks are built in layers to ease maintenance-related issues.
Multiple layers consist of different devices integrated together. The
layers of networks may consist of devices like transmitters, receivers,
media converters, network switches, etc.
These networks are designed with three tiers that facilitate strategic
installation, management, and maintenance, and so on. The strategic
design of a hierarchy network may comprise more than three layers,
however, the base foundation of this network consists of three layers
i.e.; core layer, distribution layer, and access layer.
A core switch is the network switch installed at the backbone
of the layered or hierarchy network. These data switches are
responsible for routing and data switching at the core layer of
the network. The data routed and switched by the core switch
is carried forward to the bottom layers of the network such as
the distribution and access layer. This means the performance
of the entire network relies on the data routed and switched
by the core switch. Generally, multiple data switches are used
at the core layer of a network so that a large amount of data
Core can be routed to the layers in the hierarchy.
Another reason for using multiple data switches at the core
Switches layer is to prevent the crowding of data packets. If data
packets are highly crowded at distribution and access layers,
the backpressure of data may cause a malfunction of the core
layer.
Backpressure refers to the build up of data at an I/O switch
when buffers are full and not able to receive additional data.
No additional data packets are transferred until the bottleneck
of data has been eliminated or the buffer has been emptied.
That is why a selection of high-capacity core switches is
essential in hierarchy Ethernet networks.
Aggregation
• This layer provides the Layer 2 and 3
demarcation for all northbound and
southbound traffic, and it processes most
of the eastbound and westbound traffic
within the data center.
Distribution/Aggregation
• These switches bridge the core layer and access layer. The main
responsibility of these switches is to ensure the routing of data
to correct devices in the access layer. However, the distribution
layer handles maximum data traffic as the data packets are
pushed through the core layer to the distribution layer. The
distribution switches categorize these data packets and transmit
them to specific workgroups.
Access
• This tier is the point at which the servers
physically attach to the network. The server
components consist of different types of
servers:
• Blade servers with integral switches
• The access-layer network infrastructure also
consists of various modular switches and
integral blade server switches. Switches
provide Layer 2 topologies, fulfilling the
various reqs
Multi-tenancy cost
Characteristics Elastic resources Increase, dec
Move
workload
Flexible service Additional vm
management Server
virtualization
Applications
Cloud
ISPs IXPs
computing
University
Research
campus
Issues in Datacenter
Scalability Reliability Congestion
Cooling Power Traffic
system consumption engineering
Availaibility vs. Reliability
• Availability measures the ability of a piece of equipment to be
operated if needed
• Reliability measures the ability of a piece of equipment to perform its
intended function for a specific interval without failure.
• Availability refers to the percentage of time that the infrastructure,
system, or solution remains operational under normal circumstances
in order to serve its intended purpose
• Reliability can be used to understand how well the service will be
available in context of different real-world conditions.
Availability
• Data Centers are measured in 4 tiers:
• Tier 1: No more than 29 hours of potential service interruption in a calendar year
(99.671% uptime).
• Tier 2: No more than 22 hours (99.741%).
• Tier 3: No more than 1.6 hours (99.982%).
• Tier 4 no more than 26.3 minutes (99.995%).
Assignment # 01
• Submission Deadline: Oct 9th, 2024 till 12:00 am
• To be submitted individually as pdf format file.
• To be uploaded on Google drive link:
• https://drive.google.com/drive/folders/1OZO2WZBDnfRQGRejP4leD16
KYShVOgde?usp=drive_link (BEIS – 02)
• https://drive.google.com/drive/folders/1bMwLmCqU_uZBAoGjBdymD
b-oFJ9EftsR?usp=drive_link (BEIS – 03)
• 10% deduction per day on late submission