Software Defined Networking

Traditional Computing Vs Modern Computing

Big Data: Fueling Smarter Commerce Cycle
Safety / Security
High
Exa

Sophistication of Analysis
Data Volume

Computational Needs
10s millions cameras

Healthcare
Peta Video

Med
Tera Image
1B medical images/yr

Customer Audio
Giga Text
Structured data
1B camera phones
1990’s 2000’s 2010’s 2020’s Low

Media Wide Area Imagery Digital Marketing Enterprise Video

72 video hrs/minute 100’s TB per day 10+% of video views Used by 1/3 of enterprises
Lack ofAutomated, Programmable Network
Today: VM on-boarding
is measured in minutes But… today:
93%
multi-tier virtual-system
60+% connectivity is measured
in days.
John Manville, Cisco IT; The
Power of a Programmable
Cloud, OFC 2013 (OM2D.2): “It

? takes about 5 days from an end-

end point of view to provision
something like that (a multi-tier
Use storage Use network system).” Goal is to “get at least
Percent of servers virtualization virtualization
virtualized
to sub-one day.”
 Limitations of Current Networks

• Enterprise networks are difficult to manage

• “New control requirements have arisen”:

• Greater scale
• Migration of VMS

• How to easily configure huge networks?

5
What is SDN?

SDN Definition SDN Benefits

Centralization of control of the Efficiency: optimize existing

network via the applications, services, and
infrastructure
Separation of control logic to off-
device compute, that
Scale: rapidly grow existing
applications and services
Enables automation and
orchestration of network services via
Innovation: create and deliver new
types of applications and services and
Open programmatic interfaces
business models
Source: Adopted from SDN Central (Software-Defined Networking (SDN) Use Cases) 6
SDN Innovation & Components
Well-defined
Open API
App App App App

SDN Controller/ Network Operating System

OpenFlow

OpenFlow compliant OS

Packet-Forwarding Hardware

OpenFlow compliant OS OpenFlow compliant OS

Packet-Forwarding Hardware Packet-Forwarding Hardware

SDN Approach
Software-defined Networking (SDN)
It is an approach to computer networking that allows network administrators to
programmatically initialize, control, change, and manage network behavior
dynamically via:

 open interfaces
 abstraction of lower-level functionality

SDN is meant to address the fact that the static architecture of traditional
networks doesn't support the dynamic, scalable computing and storage needs of
more modern computing environments such as data centers.
This is done by decoupling or disassociating the system that makes decisions
about where traffic is sent (the SDN controller, or control plane) from the
underlying systems that forward traffic to the selected destination (the data
plane).
Traditional Networks
The SDN architecture is:
 Directly programmable: Network control is directly programmable because it is decoupled
from forwarding functions.
 Agile: Abstracting control from forwarding lets administrators dynamically adjust network-
wide traffic flow to meet changing needs.
 Centrally managed: Network intelligence is (logically) centralized in software-based SDN
controllers that maintain a global view of the network, which appears to applications and
policy engines as a single, logical switch.
 Programmatically configured: SDN lets network managers configure, manage, secure, and
optimize network resources very quickly via dynamic, automated SDN programs, which
they can write themselves because the programs do not depend on proprietary software.
 Open standards-based and vendor-neutral: When implemented through open standards, SDN
simplifies network design and operation because instructions are provided by SDN
controllers instead of multiple, vendor-specific devices and protocols.
You win with Software Defined Networking
SIMPLIFICATION
Network Simplification

Simple Agile Automated

Lower TCO Availability & Scale Faster Services

Traditional vs. SDN networks
Point of Difference Traditional Networks Software-Defined Network
Today’s business environment expects zero service By placing the control logic outside of the network
disruption — this means networks must dynamically adapt hardware, businesses have more flexibility to
Dynamics to changing traffic by anticipating user demands. With control programmability, automation etc. This
traditional networks becoming increasingly complex, it is helps them develop scalable networks that adapt to
an uphill task to match market demands. changing business needs in quick time.
Implementing a network-wide policy requires configuring SDN by contrast allows network operators to
Application of at the device-level, making it difficult to apply a consistent programmatically configure a simplified network
set of access, security, QoS, and other policies in today’s abstraction ensuring higher chances of a consistent
Policies and mobile environment. This leaves the enterprise open to application of policies, security etc. across the
Security security breaches, non-compliance with regulations, and so entire network.
on.
It is a challenge for the network to keep pace with growing With SDN, since it is possible to abstract the
demands on the data center. Typically, link oversubscription underlying infrastructure, network manageability,
Scalability has allowed scaling of the network by estimating traffic scalability, and agility can be enhanced.
patterns – this however, is not a reliable method anymore.
In the current scenario, enterprises are constrained by SDN control software is vendor-agnostic and can
vendors equipment product cycles and hence are unable to control any network device. To ensure faster
Control of Network respond quickly to business demands. Lack of standard and responses to changes, SDN-based orchestration
Devices open interfaces hinders customizing the network to and management tools can help quickly deploy,
individual environments. configure, and update devices across the entire
network.
Two possible views on SDN:
OpenFlow, SDN, NFV Evolution
What is Open Flow
 OpenFlow is like an x86 instruction set for the network nodes.

 Provides open interface to “black box” networking node (ie. Routers, L2/L3
switch) to enable visibility and openness in network

 Separation of control plane and data plane.

 The datapath of an OpenFlow Switch consists of a Flow Table, and an action
associated with each flow entry
 The control path consists of a controller which programs the flow entry in the flow
table

20
Traditional Switch Forwarding
Open Flow Switch Forwarding
Open Flow Illustration
OpenFlow Controller
 Manages one or more switch via OpenFlow channels.
 Uses OpenFlow protocol to communicate with a OpenFlow aware switch.
 Acts similar to control plane of traditional switch.
 Provides a network wide abstraction for the applications on north bound.
 Responsible for programming various tables in the OpenFlow Switch.
 Single switch can be managed by more than one controller for load balancing
or redundancy purpose. In this case the controller can take any one of the
following roles.
 Master.
 Slave.
 Equal.

24
Components of OpenFlow Network
Controller
OpenFlow Switch specification

OpenFlow Switch PC

sw Secure
Channel

hw Flow
Table
OpenFlow Channel
 Used to exchange OpenFlow message between switch and controller.
 Switch can establish single or multiple connections to same or
different controllers (auxiliary connections).

 A controller configures and manages the switch, receives events from

the switch, and send packets out the switch via this interface

 The SC connection is a TLS/TCP connection. Switch and controller

mutually authenticate by exchanging certificates signed by a site-
specific private key

26
OpenFlow Switch
 Consists of one or more flow tables, group table and
meter table.
 A single switch can be managed by one or more
controllers.
 The flow tables and group table are used during the lookup
or forwarding phase in order to forward the packet to
appropriate port.
 Meter table is used to perform simple QOS operations like
rate-limiting to complex QOS operations like DiffServ etc
27
Pipeline Processing

OpenFlow Switch Specification

Open Flow
 General Myth
 SDN is Open Flow

 Reality
 OpenFlow is an open API that provides a standard interface
for programming the data plane switches

30
SDN Use Cases
• Resource Managements
• Energy conservation, routing, and management in data centers
• Seamless use of diverse wireless networks
• Network based load balancing
• Traffic engineering
• Slicing and scalable remote control/management of home networks
• Experimentation with new approaches and protocols using selected
production traffic
• Run virtual shadow network for traffic analysis and re-configuration
• And many more …
Simulators Used for SDN
NS3: It supports OpenFlow switches which are restricted to be simulated
only.

MININET: Network emulation software that allows you to launch a virtual

network with switches, hosts and an SDN controller all with a single command.
Mininet supports research, development, learning, prototyping, testing,
debugging, and any other tasks that could benefit from having a complete
experimental network on a laptop or other PC.
Adaptive Resource Management

• Two approaches have been implemented to improve the performance of several SDN

topologies by setting different values of capacities for each link adaptively.

• These approaches are:

 Statistical approach based on MMSE.

 Genetic Algorithm.

• The results are shown in next slides.

 Performance Evaluation of the Proposed Solution

Statistical Approach Genetic Algorithm

Video Performance before and after Implementing Proposed Solution