Mod 3 end copy

Enterprise Network Design
Module 3
Lecture 1
Designing Basic Campus and Data Center
Networks
St. Francis Institute of Technology
Department of Information Technology 1
Dr. Minal Lopes9th September 2020
The material in this presentation belongs to St. Francis Institute of Technology and is solely for educational purposes. Distribution and modifications of the content is prohibited.
Ref: Authorized Self-Study Guide, Designing for Cisco Internetwork Solutions (DESGN), Second Edition, Cisco
Press-Diane Teare, Chapter 4 Designing Basic Campus and Data Center Networks

Designing an Enterprise Campus
 Enterprise Campus network is the foundation for
businesses - enhancing productivity, and providing a
multitude of services to end users.
 The following characteristics are considered when
designing the campus network:
 Network application characteristics
 Environmental characteristics
 Infrastructure device characteristics

 Network application characteristics:
 The organizational requirements, services and
applications place stringent requirements on a campus
network solution
 For example, bandwidth and delay can be different for
different services and applications

 Environmental characteristics: The network’s
environment includes its geography and the transmission
media used
 The physical environment of the building
 the number of network nodes (end users, hosts, network
devices)
 distribution of network nodes
 distance between the network nodes
 Other factors include space, power, heating, ventilation, air
conditioning support for the network devices
 Cabling is one of the biggest long-term investments in network
deployment
 transmission media selection should depend on the emerging
technologies that might be deployed over the same
infrastructure in the future

 Infrastructure device characteristics:
 The characteristics of the network devices influence the
design (ex: they determine the network’s flexibility) and
contribute to the overall delay
 Trade-offs between data link layer switching and multilayer
switching , based on network layer addresses, transport
layer and application awareness needs to be considered
 High availability and high throughput are requirements that
might require consideration throughout the infrastructure
 Most Enterprise Campus designs use a combination of data
link layer switching in the access layer and multilayer
switching in the distribution and core layers

Network Application Characteristics and
Considerations
 The network application’s characteristics and
requirements influence the design in many ways
 The network demands of applications that are critical to
the organization, determine traffic patterns inside the
Enterprise Campus network
 This influences bandwidth usage, response times and the
selection of the transmission medium
 Different types of application communication that can
generate different traffic patterns are
 Peer-Peer
 Client–local server
 Client–Server Farm
 Client–Enterprise Edge server

Considerations
 Peer-Peer Applications
 From the network designer’s perspective, peer-peer
applications include applications in which the majority of
network traffic passes from one network device to another
through the organization’s network
 Typical peer-peer applications:
 Instant messaging
 IP phone calling
 File sharing
 videoconferencing

Considerations
 Client–Local Server Applications
 clients and servers are attached to
a network device on the same LAN
segment and follow the 80/20
workgroup rule for client/server
applications
 This rule indicates that 80
percent of the traffic is local to
the LAN segment and 20 percent
leaves the segment
 With increased traffic on the
corporate network an organization
might split the network into
several isolated segments

Considerations
 Client–Server Farm Applications
 In a large organization, application
traffic might have to pass across
more than one LAN or VLAN to
reach servers in a Server Farm.
 Client–Server Farm applications
apply the 20/80 rule, where only
20 percent of the traffic remains
on the local LAN segment and 80
percent leaves the segment to
reach centralized servers, the
Internet and so on.

Considerations
 Client–Enterprise Edge
Applications
 Typical Enterprise Edge
applications are based on
web technologies.
 Examples: external mail,
DNS servers, public web
servers

Considerations
 Client–Enterprise Edge
Applications
 Typical Enterprise Edge
applications are based on web
technologies.
 Examples: external mail, DNS
servers, public web servers

Considerations
 Application Requirements comparison w.r.t. network
parameters
Parameter Peer-to-Peer
Client-
Local
Servers
Client-
Server
Farm
Client-
Enterprise
Edge Servers
Connectivity type Switched Switched Switched Switched
Total required
throughput
Medium to
high
Medium High Medium
High availability Low to high Medium High High
Total network costs
Low to
medium
Medium High Medium

Considerations
 Connectivity
 The wide use of Layer 2 switching has revolutionized local-
area networking
 It provides increased performance and more bandwidth for
satisfying the requirements of new organizational
applications
 LAN switches provide this performance benefit by increasing
bandwidth and throughput for workgroups and local servers

Considerations
 Throughput
 The required throughput varies from application to
application
 An application that exchanges data between users in the
workgroup usually does not require a high throughput
network infrastructure
 However, organizational-level applications usually require a
high-capacity link to the servers, which are usually located in
the Server Farm

Considerations
 High availability
 The high availability of an application is a function of the
application and the entire network between a client
workstation and a server located in the network.
 the network design primarily determines the network’s
availability
 Redundancy in the Building Distribution and Campus Core
layers is recommended

Considerations
 Total network cost
 Depending on the application and the resulting network
infrastructure, the cost varies from low in a peer-peer
environment to high in a network with redundancy in the
Building Distribution, Campus Core and Server Farm
 In addition to the cost of duplicate components for
redundancy, costs include the cables, routers, switches,
software and so forth

Module 3
Lecture 2
Networks
Dr. Minal Lopes10th(beitb), 12th (beita) September 2020

Campus Network Design Considerations
Network application
characteristics
Peer-Peer
Client–local server
Client–Server Farm
Client–Enterprise
Edge server
Environmental
characteristics
Network Geography
Considerations
Intrabuilding
Interbuilding
Distant remote
building
Transmission Media
Considerations
Copper
Fiber
Wireless
Infrastructure device
characteristics

Environmental Characteristics and
Considerations
 Network Geography Considerations
 The location of Enterprise Campus nodes and the distances
between them determine the network’s geography
 Nodes, including end-user workstations and servers, can be
located in one or multiple buildings
 Based on the location of nodes and the distance between
them, the network designer decides which technology should
interconnect them based on the required maximum speed,
distance etc
 Following structures are widely used in network geography:
 Intrabuilding
 Interbuilding
 Distant remote building

Considerations
 Intrabuilding Structure
 This network structure provides connectivity for all
end nodes located in the same building and gives
them access to the network resources
 The Building Access and Building Distribution layers
are typically located in the same building
 User workstations are usually attached to the
Building Access switches in the floor wiring closet
with twisted-pair copper cables
 Wireless LANs can also be used to provide
intrabuilding connectivity without the limitations of
wires or cables
 Access layer switches usually connect to the Building
Distribution switches over optical fiber, provide
better transmission performance and less sensitivity
to environmental disturbances than copper.
 Depending on the connectivity requirements to
resources in other parts of the campus, the Building
Distribution switches may be connected to Campus
Core switches.

Considerations
 Interbuilding Structure
 This network structure provides
connectivity between the
individual campus buildings’
central switches (in the Building
Distribution and/or Campus
Core layers)
 These buildings are usually in
close proximity, typically only a
few hundred meters to a few
kilometers apart
 Nodes in campus buildings,
share common devices such as
servers, thus demand for high-
speed connectivity between the
buildings is high.

Distant Remote Building Structure
 Interbuilding Structure
 Within a campus, companies
might deploy their own
physical transmission media.
 To provide high throughput
without excessive
interference from
environmental conditions,
optical fiber is the medium of
choice between the buildings
 The Building Distribution
switches might be connected
to Campus Core switches

Considerations
 Distant Remote Building Structure

Considerations
 Distant Remote Building Structure
 For remote building connectivity (within a metropolitan area),
choice of physical media is most important factor
 The speed and cost of the network infrastructure depend
heavily on the media selection
 The network designer must identify the organization’s critical
applications and then select the intelligent network services
such as QoS for optimal use of bandwidth
 Some companies might own their media (fiber, microwave,
copper lines).
 If the organization does not own physical transmission media,
the Enterprise Campus must connect through the Enterprise
Edge using connectivity options from public service providers
(traditional WAN links or Metro Ethernet)

Transmission Media Considerations
 The type of cable is an important consideration when deploying a
new network or upgrading an existing one
 Cabling infrastructure represents a long-term investment—usually
installed to last for ten years or more
 Factors to consider,
 The cost of the medium (including installation costs)
 the available budget
 the technical characteristics - signal attenuation and
electromagnetic interference
 Common mediums:
 Twisted-pair cables (copper)
 optical cables (fiber)
 wireless (satellite, microwave and IEEE 802.11 LANs)

 Copper:
 Twisted-pair cables consist of four pairs of
isolated wires that are wrapped together
in plastic cable.
 With unshielded twisted-pair (UTP), no
additional foil or wire is wrapped around
the core wires. This makes these wires less
expensive, but also less immune to
external electromagnetic influences than
shielded twisted-pair cables.
 Twisted-pair cabling is widely used to
interconnect workstations, servers, or
other devices from their network interface
card (NIC) to the network connector at a
wall outlet.

 Copper:
 Category 5 or greater are used for speeds
of 100 Mbps or higher.
 Category 6 is recommended for gigabit
Ethernet.
 Because of the possibility of signal
attenuation in the wires, the maximum
cable length is usually limited to 100
meters.
 Collision detection needs to be taken care

 Optical Fiber:
 For transmission mediums with
distances longer than 100 meters and
immunity to electromagnetic
interference
 Types of optical cable:
 multimode (MM)
 used for relatively short distances.
 LEDs are used as source
 single-mode (SM)
 Lasers are used as source
 limits dispersion and loss of light
 more expensive than multimode fiber

 Wireless:
 In-building WLAN equipment includes access
points (AP) that perform functions similar to
wired networking hubs and PC client adapters.
 APs are distributed throughout a building to
expand range and functionality for wireless
clients
 IEEE 802.11g allow speeds of up to 54 Mbps in
the 2.4-GHz band over a range of about 100 feet
 IEEE 802.11b standard supports speeds of up to
11 Mbps in the 2.4- GHz band
 The IEEE 802.11a standard supports speeds of
up to 54 Mbps in the 5-GHz band

Transmission Media Type Characteristics

Cabling example in Campus Network

Module 3
Lecture 3
Networks
Dr. Minal Lopes12th (beitb), 18th (beita) September 2020

Campus Network Design Considerations
Network application
characteristics
Peer-Peer
Client–local server
Client–Server Farm
Client–Enterprise
Edge server
Environmental
characteristics
Network Geography
Considerations
Intrabuilding
Interbuilding
Distant remote
building
Transmission Media
Considerations
Copper
Fiber
Wireless
Infrastructure device
characteristics
Convergence time
IP Multicasting
QoS

Infrastructure Device Considerations
 Network end-user devices are commonly connected
using switched technology rather than using a shared
media segment.
 Switched technology provides dedicated network
bandwidth for each device on the network. Switched
networks can support network infrastructure services,
such as QoS, security and management; a shared media
segment cannot support these features.
 Deciding whether to deploy pure data link layer
switches or multilayer switches in the enterprise
network requires a full understanding of the network
topology and user demands

 Factors that determine which switch to use:
 Infrastructure service capabilities: The network services
that the organization requires (IP multicast, QoS, etc)
 Size of the network segments: How the network is
segmented and how many end devices will be connected
 Convergence time: The maximum amount of time the
network will be unavailable in the event of network outages.
 Cost: The budget for the network infrastructure. (multilayer
switches are typically more expensive than their Layer 2
counterparts; however, multilayer functionality can be
obtained by adding cards and software to a modular Layer 2
switch)

factors affecting choice of switch
 Convergence Time
 Networks can use layer 2 and/or layer 3 switches
 If layer 2 switches are used, STP protocol is used to avoid loops
 STP takes 30-50 sec to converge; its slow
 Core connectivity: To avoid STP convergence problem, use routed
links instead of VLAN trunks
 Core connectivity: use EIGRP, OSPF on routed links
 convergence is within seconds because all the devices detect their
connected link failure immediately and send respective routing
updates
 In a mixed Layer 2 and Layer 3 environment, the convergence time
depends not only on the Layer 3 factors (including routing
protocol timers such as hold-time and neighbour loss detection),
but also on the STP convergence.

 IP Multicast
 A traditional IP network is not efficient when sending the
same data to many locations
 IP multicast technology enables networks to send data to a
group of destinations in the most efficient way
 Multicast groups are identified by Class D IP addresses,
which are in the range from 224.0.0.0 to 239.255.255.255
 IP multicast involves new protocols for network devices,
 IGMP: for informing network devices which hosts require
which multicast data stream
 PIMRP: for determining the best way to route multicast traffic.
It is independent of EIGRP or OSPF
IGMP: Internet Group Management Protocol
PIMRP: Protocol Independent Multicast Routing Protocol

 QoS:
 A campus may contain bandwidth intensive or delay
sensitive network applications
 Demands secure network, predictable delays, guaranteed
services
 Such services need QoS implementation
 QoS in Access, distribution and core layers
 QoS Mechanisms:
 Classification: the process of partitioning traffic into multiple
priority levels or classes of service. Information in the frame or
packet header is inspected and the frame’s priority is
determined
 Marking: the process of changing the priority or class of
service setting within a frame or packet to indicate its
classification

 QoS Mechanisms:
 Congestion management: Queuing
 Queuing separates traffic into various queues or buffers
 the marking in the frame or packet can be used to determine
which queue traffic goes in
 When a network interface is often congested, queuing
techniques ensure that traffic from the critical applications is
forwarded appropriately
 For example, real-time applications such as VoIP and stock
trading need to be forwarded with the least latency and jitter
 Congestion Management: Scheduling
 Scheduling is the process that determines the order in which
queues are serviced.

 QoS Mechanisms:
 Policing and shaping: Policing and shaping tools identify
traffic that violates some threshold level and reduces a
stream of data to a predetermined rate or level
 Traffic shaping buffers the frames for a short time.
 Policing simply drops or lowers the priority of the frame

QoS in Campus networks

Enterprise Campus Design
 Functional Modules:
 Campus Infrastructure
 The Building Access layer
 The Building Distribution layer
 The Campus Core layer
 Server Farm
 Edge Distribution (optional)

Enterprise Campus Requirements
 Enterprise Campus module has different requirements
 Technology High Availability Cost per port
 Scalability Performance

Module 3
Lecture 4
Networks
Dr. Minal Lopes17th (beitb), 21st (beita) September 2020

Enterprise Campus Design
 Functional Modules:
 Campus Infrastructure
 The Building Access layer
 The Building Distribution layer
 The Campus Core layer
 Edge Distribution (optional)
 Server Farm

Building Access Layer Design Considerations
 Requirements:
 Current cabling and cabling options
 Link capacity of building distribution layer switches
 Multicast traffic management
 Number of end users at present and in future
 No. of ports available
 No. of VLANs per access switch
 QoS
 Redundancy level
 Security

Building Access Layer Design Considerations
 Some best practices:
 VLAN and spanning tree protocol Management
 VLAN should be limited to a single wiring closet
 For highly reliable, predictable network, avoid using STP
 If STP is unavoidable, use layer 3 routing to isolate different
STP domains
 Trunks Management
 To support DTP encapsulation, set DTP to ‘desirable’ and ‘desirable
with negotiate option’ on one side each.
 To avoid broadcast propagation, manually prune the VLANs that
are not used from trunked interfaces
 To decrease the potential for operational error, use the VTP
transparent mode.
 To prevent VLAN hopping, identify the ports on which hosts will be
attached and disable trunking on such ports

Building Distribution Layer Design Considerations
 An intermediate layer between, access and core layers
 It aggregates data from access layer and provides it to the high
speed campus core layer
 Uses multilayer switches
 Incorporates services such as, security, high availability, traffic
filtering, QoS etc.
 Requirements:
 Current cabling and cabling options
 No. of building distribution switches
 Number of devices per distribution switch
 Redundancy type and level
 No. of redundant switches
 No. of redundant links, no. of uplinks, speed of uplinks

 First-Hop Redundancy Protocols
 Layer 2 or layer 3 switching is used between access and
distribution layer
 If layer 2 is used, convergence time on node or link failure
depends on default gateway redundancy and failover time
 Default gateway redundancy is provided by building
distribution switches by using first hop redundancy protocols
 First hop redundancy protocols:
 Gateway load balancing protocol (GLBP)
 Hot standby router protocol (HSRP)
 Virtual router redundancy protocol (VRRP)

 Layer 3 routing protocols between distribution & core
 For fast convergence on the failure of links, deployment of
routing protocols is advised between distribution and core
switches
 This helps in rerouting network traffic to alternate equal cost
path
 For predictable convergence, prefer building triangles instead
of rectangles for alternate equal cost paths.

 Layer 3 routing protocols between distribution & core
 In triangle redundancy, alternate paths are of equal cost
 Entries for these paths are stored in a routing table
 On any link failure, no recalculation for routing path is required
 An alternate route is instantly chosen from the routing table
 In rectangle redundancy, on link failure, recalculation of an
alternate route is required, increasing convergence time

Building Core Layer Design Considerations
 For small or medium campus, core layer is merged with
distribution layer
 Separate core layer design is used only for larger campus
networks
 Core layer gets traffic from distribution layer
 Integrate with server farm and enterprise edge
 It uses multilayer switches

 Requirements:
 Enterprise Edge and WAN connectivity
 LAN, WAN, MAN convergence
 High availability, redundancy and scalability requirements
 Performance requirements
 No. of high capacity ports
 No. of intelligent multilayer switches

 Large campus Design
 Core layer should have more than
one multi layer switch
 This ensures, high availability,
scalability and redundancy
 Figure shows, multilayer switches
in core and distribution
 Triangle redundancy is used

 Advantages:
 The network can be scaled to large size
 Supports load sharing
 No STP required
 Provides multicast and broadcast control
 Supports network infrastructure services
 Fast recovery from link failure
 High throughput

 Small Campus Design
 No of workstations, network
devices, servers are less (app.200)
 Uses single wiring closet
 Scaling, high speed switching is not
required
 Therefore, core layer can be
combined with distribution layer

 Medium Campus Design
 No of workstations, network
devices, servers are appr. 200-
1000
 Servers are connected to
distribution switches
 Redundancy is added for high
performance and availability
 core layer can still be combined
with distribution layer

Edge Distribution Design
 Edge Distribution is used
to connect Enterprise
Edge module to Campus
Core
 Additional filtering,
security mechanisms are
required
 Provide last line of defense
by protecting against,
 Unauthorized access
 IP spoofing
 Packet sniffing

Edge Distribution Design
 Functionalities of Edge distribution switches:
 Aggregate Enterprise Edge connectivity
 Filter and route traffic to the campus core layer
 Provide security to devices in Edge Module
 Restrict unauthorized access
 Avoid packet sniffing
 Filter traffic

Server Placement
 Servers in a network can be located at various places:
 Building access layer
 Building distribution layer
 Campus core layer
 Server farm
 Data center

Server Placement
 Servers in Building access Layer
 Its mostly used in small campuses
 Servers are directly connected to building access switches
 they mostly get traffic local to a workgroup that corresponds
to a VLAN

Server Placement
 Servers in Building Distribution Layer
 Its mostly used in medium campus networks
 Servers are directly connected to building distribution
switches
 they can get traffic local to a workgroup as well as from clients
placed at different VLANs.

Server Placement
 Servers in Campus Core
 Campus core may have
separate core switches, in
large campus networks
 In medium sized
networks, servers are
directly connected to
distribution switches
 Due to limited ports,
security and QoS services
are added in distribution
layer

Server Placement
 Servers in Server Farm
 It contains multilayer and
layer 2 switches
 High speed connectivity
(Gigabit Ethernet) is used to
connect servers to server farm
 ACLs are implemented at
distribution layer
 Redundancy and redundancy
protocols are implemented at
distribution layer
 Communication amongst the
servers is handled by server
farm

Module 3
Lecture 4
Networks
Dr. Minal Lopes21st (beitb), 22nd (beita) September 2020

Enterprise Data Center Design Considerations
 What is a Data Center?
 It’s a centralized repository of a large group of networked
servers used for storage, processing, management and
dissemination of business data.
 An Enterprise data center is a custom shelf of enterprise
applications.
 It facilitates load sharing, high availability, redundancy and
reliability for business critical applications.
 It is supported by proper heating, ventilation, air conditioning
(HVAC), power supply, security system and surveillance
system.
 Modern data center designs offer flexibility, scalability and
reduced maintenance

 Data Center Evolution
 Historically, most Enterprise Data Centers grew rapidly as
organizational requirements expanded.
 Applications were implemented as needed, often resulting in
underutilized, isolated infrastructure silos.
 Each silo was designed based on the specific application being
deployed
 So a typical data center supported a broad assortment of
operating systems, computing platforms, and storage systems.
 This resulted in various application “islands” that were
difficult to change or expand and expensive to manage,
integrate, secure and back up.

 This server-centric data center model is now evolving to a
service-centric model
 Deployed with virtual machine software: Vmware, Xen
 This broke the one-to-one relationship between applications
and the server hardware and operating system on which they
run.
 Virtual machine software allows multiple applications to run
on a single server, independent of each other and of the
underlying operating system.

 Removal of storage from the server:
 consolidating storage in pools
 Networked storage (storage area networks [SAN]) allows
easier management, provisioning, improved utilization, and
consistent recovery practices

 The creation of pools of one-way, two-way or four-way servers
that can be pooled and provisioned on demand.
 One-way servers have a single processor, two-way servers
have two processors, and four-way servers have four
processors.
 The consolidation of I/O resources so that the I/O can be pooled
and provisioned on demand for connectivity to other servers,
storage and LAN pools.

Enterprise Data Center Infrastructure
 Three layers:
 Core
 Aggregation(Distribution)
 Access
 Access Layer
 Provide connectivity for
layer 2, layer 3 and main
frame
 Use high performance,
low latency layer 2
switches
 Servers can be single or
dual attached

Dual Attached
Server
Blade Server
Blade Chasis
NIC Teaming

 Aggregation(Distribution) Layer
 Layer 3 connectivity is used to avoid STP
 Incorporates security services such as Firewalls, IDS
 This makes easy sharing of security services across all the
servers
 This also reduces total cost of ownership and complexity and
manage security services
 Core Layer
 Switch pair should support campus core and DC aggregation
 Provision of 10 Gigabit Ethernet ports for connectivity
 Provision of campus distribution/core layer isolation from DC
core

Mod 3 end copy

Recommended

Recommended

More Related Content

What's hot

What's hot (18)

Similar to Mod 3 end copy

Similar to Mod 3 end copy (20)

Recently uploaded

Recently uploaded (20)

Mod 3 end copy