1.

Introduction to OSPF

Open Shortest Path First (OSPF) --- includes single-area and multiarea

OSPFv2 used for IPv4 networks.

OSPFv3 used for IPv6 networks

OSPF is a link-state routing protocol that uses the concept of areas

RIP using hop count does not scale well in larger networks with multiple paths of varying speeds

OSPF offers faster convergence and scales to much larger network implementations.

A network administrator can divide the routing domain into distinct areas that help control routing
update traffic.

A link is an interface on a router

A link is also a network segment that connects two routers, or a stub network such as an
Ethernet LAN that is connected to a single router

Information about the state of a link is known as a link-state. link-state information includes
network prefix, prefix length, and cost.

2. Components of OSPF

routing protocol messages to exchange route information. The messages help build data
structures, which are then processed using a routing algorithm.

2.1. Routing protocol messages:

Routers running OSPF exchange messages to convey(carry) routing information using five
types of packets. These packets are used to discover neighboring routers and also to exchange
routing information to maintain(continue) accurate(precise) information about the network.

1) Hello packet

2) Database description packet

3) Link-state request packet

4) Link-state update packet

5) Link-state acknowledgment packet

2.2. Data Structures

OSPF messages are used to create and maintain three OSPF databases, as follows:
Database Table/creates Description

Adjacency Database Neighbor Table ● List of all neighbor routers to which a

router has established bidirectional
communication.
● This table is unique for each router.
● Can be viewed using the show ip
ospf neighbor command.

Link-state Database Topology Table ● Lists information about all other

routers in the network.
● This database represents the network
topology.
● All routers within an area have
identical LSDB.
● Can be viewed using the show ip
ospf database command.

Forwarding Database Routing Table ● List of routes generated when an

algorithm is run on the link-state
database.
● The routing table of each router is
unique and contains information on
how and where to send packets to
other routers.
● Can be viewed using the show ip
route command.

The tables are kept and maintained in RAM

2.3. Algorithm

topology table using the Dijkstra shortest-path first (SPF) algorithm.

The SPF algorithm is based on the cumulative cost to reach(arrive) a destination

The SPF algorithm creates an SPF tree → calculating the shortest path to each node.

The SPF tree → used to calculate → best routes.

OSPF places(put) the best routes into the forwarding database, which is used to make the
routing table.

3. Link-State Operation

To maintain routing information, OSPF routers complete a generic link-state routing process to
reach(alcanzar) a state of convergence. Each link between routers is labeled with a cost value. In
OSPF, cost is used to determine the best path to the destination. The following are the link-state
routing steps that are completed by a router:
 1. Establish Neighbor Adjacencies
2. Exchange Link-State Advertisements
3. Build the Link State Database
4. Execute the SPF Algorithm
5. Choose the Best Route

EIGRP (Enhanced Interior Gateway Routing Protocol)

EIGRP supports classless routing and VLSM, route summarization, incremental updates, load
balancing and many other useful features. It is a Cisco propriatery protocol

Routers running EIGRP must become neighbors before exchanging routing information. To
establish neighbor relationships, routers send hello packets every couple of seconds.
Hello packets are sent to the multicast address of 224.0.0.10.

Note: On LAN interfaces hellos are sent every 5 seconds. On WAN interfaces every
60 seconds

Each EIGRP router stores routing and topology information in three tables:

● Neighbor table – stores information about EIGRP neighbors

● Topology table – stores routing information learned from neighboring routers
● Routing table – stores the best routes

EIGRP uses Reliable Transport Protocol (RTP) for sending messages.

EIGRP calculates its metric by using bandwidth, delay, reliability and load. By default, only
bandwidth and delay are used when calulating metric, while reliability and load are set to zero.

EIGPR uses the concept of autonomous systems. An autonomous system is a set of EIGRP
enabled routers that should become EIGRP neighbors. Each router inside an autonomous
system must have the same autonomous system number configured, otherwise routers will not
become neighbors.

The following fields in a hello packet must be the identical in order for routers to
become neighbors:
● ASN (autonomous system number)
● subnet number
● K values (components of metric)

Routers consider the neighbor to be down after a hold-down timer has expired. Hold-
down timer is, by default, three times the hello interval. On the LAN network the hold-
down timer is 15 seconds.
Two terms that you will often encounter when working with EIGRP are feasible and
reported distance. Let’s clarify these terms:

● Feasible distance (FD) – the metric of the best route to reach a network.
That route will be listed in the routing table.
● Reported distance (RD) – the metric advertised by a neighboring router for a
specific route. In other words, it is the metric of the route used by the
neighboring router to reach the network.

EIGRP has been configured on R1 and R2. R2 is directly connected to the subnet
10.0.1.0/24 and advertises that subnet into EIGRP. Let’s say that R2’s metric to
reach that subnet is 28160. When the subnet is advertised to R1, R2 informs R1 that
its metric to reach 10.0.1.0/24 is 28160. From the R1’s perspective that metric is
considered to be the reported distance for that route. R1 receives the update and
adds the metric to the neighbor to the reported distance. That metric is called the
feasible distance and is stored in R1’s routing table (30720 in our case).

Successor and feasible successor

A successor is the route with the best metric to reach a destination. That route is
stored in the routing table. A feasible successor is a backup path to reach that
same destination that can be used immediately if the successor route fails. These
backup routes are stored in the topology table.

For a route to be chosen as a feasible successor, one condition must be met:

 ● the neighbor’s advertised distance (AD) for the route must be less than the
successor’s feasible distance (FD).

R1 has two paths to reach the subnet 10.0.0.0/24. The path through R2 has the best
metric (20) and it is stored in the R1’s routing table. The other route, through R3, is a
feasible successor route, because the feasiblility condition has been met (R3’s
advertised distance of 15 is less than R1’s feasible distance of 20). R1 stores that
route in the topology table. This route can be immediately used if the primary route
fails.

Diffusing Update Algorithm (DUAL)

EIGRP Reliable Transport Protocol (RTP)

Five different packets types are used by EIGRP:

● Update – contains route information. When routing updates are sent in

response to the metric or topology changes, reliable multicasts are used. In
the event that only one router needs an update, for example when a new
neighbor is discovered, unicasts are used.
● Query – a request for specific routes that always uses the reliable multicast
method. Routers send queries when they realize they’ve lost the path to a
particular network and are looking for alternative paths.
● Reply – sent in response to a query via the unicast method. Replies can
include a specific route to the queried destination or declare that there is no
known route.
● Hello – used to discover EIGRP neighbors. It is sent via unreliable multicast
and no acknowledgment is required.
 ● Acknowledgment (ACK) – sent in response to an update and is always
unicast. ACKs are not sent reliably.

OUT — Puppet

Network Automation and Programmability

Traditional Network Management

With traditional networking, you can only manage network devices one at a time
using SSH to the command line. This process is time-consuming, labor-intensive,
and prone to human error. While this can work well in smaller networks, it does not
work well in a large enterprise and does not also scale very well.

Traditional Network Monitoring Systems (NMS) such as SolarWinds, CiscoWorks,

and Cisco Prime Infrastructure have also been available for a long time and use
protocols such as SNMP and Netflow to gather information reports on the state of
the network.

While SNMP uses MIB and OID to collects useful data and can also be used to push
configuration to devices, it does not build for real-time programmatic access and has
limited functionality.

SNMP also has security concerns and can be complex to implement and operate.

These are the main challenges of traditional Networks:

● Scalability Issues
● Management Complexity
● Slower Issue Resolution

Controller-Based Networks

A Controller-based network uses controller software or appliance(device/tool) to

automate network operations. Software-defined networking uses a controller for
network automation. It allows you to centrally manage your network device’s
configuration and enable programs to automatically configure and operate networks
through the power of Application Programming Interfaces (APIs).

API: a stateless client-server model

Intelligent network automation through a controller-based network all of the
configuration settings can be easily managed in the GUI of the controller.

Only controller-based networks decouple the control plane and the data plane.

SD-Access is Cisco’s solution for network automation. It offers a single point of

automation, orchestration, and management of networks through the Cisco DNA
Center controller. SD-Access has two key components:

● Cisco Campus Fabric Solution- it is a network virtualization overlay built on

top of the network underlay(sobre la base).
● Cisco DNA Center – SDN controller for centralized management and NMS

Note:

Controllers can also provide central aggregation points for monitoring and data
analytics for more visibility and easier troubleshooting.

DNA Center can be the “controller” or single pane of glass for deploying and
managing on-premises Cisco networks including routing, switching, Software
Defined Wireless Area Network (SD-WAN), and security devices like Cisco Identity
Services Engine (ISE) and Firepower

Function of SDN(software-defined network) controller:

● coordinating VTNs(Virtual Transport Network)
● managing the topology
● making routing decisions

Southbound API to communicate between the Controllers and Network Elements.

Network Automation Tools

Multiple methods can be used to automate network management – Python scripts,

NETCONF, RESTCONF, Ansible, Puppet, SDN, Cisco DNA Center, etc. However,
not all methods are supported by all devices so you should choose which is most
suitable for your environment.

Network configuration management tools provide an automated method to execute

and monitor changes to our networks. Some tools even offer network automation
frameworks wherein a library of commands and workflows are already provided. If
such changes do occur, it either reconfigures the device or informs the network
engineers to make the change.

There are two types of configuration management tools:

agent-based tools – need the installation of a code (agent) on the managed device,
such as Puppet, Saltstack, and Chef.
agentless tools – does not need the installation of any code (agent) on them, such
as Ansible, Saltstack SSH, and Puppet Bolt.

Cisco DNA Center

The Cisco Digital Network Architecture (DNA) Center is an SDN controller that is
used to support Cisco’s Software-Defined Access for network automation. The DNA
Center provides centralized management on your system, automates configuration
changes, and performs analysis of the devices

Cisco DNA Center Assurance is a tool for monitoring and network management. It
monitors traffic statistics, the health status of the network, and an overview of all
issues in the network. It also correlates issues and provides recommendations on
what may cause such network issues to help you assist in diagnosing and
troubleshooting.
Cisco DNA Center also enhances network visibility use that visibility to improve
performance, It enables proactive prediction of network-related and security-related
risks by using telemetry to improve the performance of the network.

It support numerous extensibility options including cross domain adapters and third -
party SDKs

Functionality es software-defined controller for automation of device and services

Network programmability

is described as a set of software tools for network engineers for provisioning,

managing, and troubleshooting networks We will discuss various methods for
network programmability, such as Git, GitHub, CI/CD, and Python.

Python: is one of the most popular programming language globally

Git: is a distributed version control system that tracks changes in source code and
files.

GitHub: is a web-based repository for code

CI/CD: means Continuous Integration and Continuous Delivery, and it is a set of
operating principles and practices that permits application development teams to
deliver code changes for network programmability more frequently and reliably.

Data Serialization Formats – JSON, YAML, and XML

Here are some of the common data serialization formats used to ease the storage
and sharing of data across the Internet.

JSON – JavaScript Object Notation

Is considered a lightweight data-interchange format that translates data structures

easier to be read and written by humans. This human-readable format has a short
syntax textual format with limited data types derived from JavaScript data formats.
JSON data does not need a separate parser since they are referenced to JavaScript
objects compared to XML.

Benefits of using JSON files:

1. Self-describing
2. Simplicity and compact
3. Ease in learning and understanding

Data representation

XML – Extensible Markup Language

its simplicity in sharing and storing data. A nested textual format that is human-
readable and can also be edited using a schema-based validation used in metadata
applications and web publishing and services transfer of data.

Advantages of using XML:

1. Separates data from HTML

2. Simplified data sharing and transport
 3. Simplified platform change
4. Data availability increase
5. Useable to create a new Internet language

YAML: YAML Ain’t Markup Language

A lightweight text human-readable format that supports comments and is really easy
to edit. Considered as a superset of JSON that supports complex data types that
enable ease in mapping native data structures. The structured data is used in
configuration settings, Apps that require MySQL self-references in relational data,
and document headers. The serialized data is so readable that even the content of
the front page is displayed in YAML.

Perks(advantages) :

1. Working with a text editor is easy

2. Supports the Unicode character set
3. Extensible and expressive
4. Portability between Programming languages

The two main types of APIs used for web service that runs over the Internet and
typically use HTTP/HTTPS are SOAP and REST.

Simple Object Access Protocol (SOAP) API

SOAP APIs are standard communication protocol systems that permit different
operating systems like Linux and Windows to communicate. It is commonly used
with HTTP transport with XML data format to access web services

communicate between applications running on different operating systems, with

different technologies and programming languages.

Representational State Transfer (REST) API

is an architecture style for designing a distributed system that gives guidelines for the
structure and organization of an API for web service. REST is protocol-independent.
It supports any transport and data format like HTTP/S transport and JavaScript
Object Notation (JSON) or XML data format.

NOTE
Unlike SOAP protocol, REST APIs do not require a new message format. The HTTP
API is CRUD (Create, Retrieve, Update, and Delete).
Encoding methods : JSON y XML

Model-Driven Programmability: NETCONF & RESTCONF

NETCONF

The NETCONF protocol enables the device to expose an entire formal Application
Programming Interface (API). Applications can use this straightforward API to send
and receive full and partial configuration data sets.

● A simple network device can be managed

● Configuration data information can be retrieved
● New configuration data can be manipulated and uploaded

The protocol messages are also in XML and the mandatory transport protocol for
NETCONF is through the Secure Shell Transport Layer Protocol (SSH).

NETCONF uses a remote procedure calls (RPC) model wherein the client inputs an
RPC in XML and forwards it to a server utilizing a secure, connection-oriented
session.

YANG
YANG is a data modeling language for NETCONF as per RFC 6020. It is used to
model configuration and state data manipulated by NETCONF, NETCONF RPCs,
and notifications.

RESTCONF
RESTCONF is a protocol defined based on HTTP used for configuring data defined
in YANG using the datastore concepts defined in the NETCONF. RESTCONF uses
HTTP methods to provide Create, Read, Update, Delete (CRUD) operations on a
conceptual datastore comprising YANG-defined data, which is compatible with a
server that administers NETCONF datastores.

Configuration Management Tools – Ansible, Chef, Puppet

Most usually used management tools to simplify network configuration management
tasks which are Ansible, Chef, and Puppet.

Ansible

It is an automation platform that is capable of deployment of applications,

configuration and monitoring management, security management, and automating
cloud provisioning and deployment.

works with python programming language and YAML data format

Ansible uses SSH as a remote communication protocol (configuration management

mechanism uses TCP port 22 by default when communicating with managed nodes)

Chef

An agent is required to be installed on network devices that the Chef is managing.

Ruby is the programming/configuration language used

The Chef Server is the device that manages all the nodes on the network, while
Chef Client is the node that the chef server is managing.
The collection of codes or files in Chef is called Cookbook. Recipe is the code
being deployed on making configuration changes, and Chef Workstation is where
users interact with the management tools for network configuration and create
codes/instructions.

Puppet

typically uses an agent-based architecture for managing the network. support both
agent and agentless architecture.

When using agentless architecture, the external agent will use SSH to communicate
with the node (network device)

The Puppet Master machine is the device that manages all the nodes on the
network, while Puppet Agent is the node

The collection of codes or files in Puppet is called Module. Manifest is the code
being deployed on making configuration changes, and Puppet Console is where
users interact with a configuration management tool to create a block of code or
instruction

CRUD is short for CREATE, READ, UPDATE and DELETE operations. Only
UPDATE operation modifies an existing table or view so it is equivalent to PUT and
PATCH in HTTP methods.
PUT: fully update (i.e. replace) an existing record
PATCH: update part of an existing record
PUT is similar to POST in that it can create resources, but it does so when there is a
defined URL wherein PUT replaces the entire resource if it exists or creates new if it
does not exist.
Unlike PUT Request, PATCH does partial update. Fields that need to be updated by
the client, only that field is updated without modifying the other field.

REpresentational State Transfer (REST) requires that a client make a request to the
server in order to retrieve or modify data on the server

GET: retrieve data POST: create data PUT: fully update (i.e. replace) an existing
record PATCH: update part of an existing record DELETE: delete records

Sylog

Emergency : The most severe and results in the system being considered
unusable(inutilizable)

In case of logging significant events, the syslog messages needs to be transported

over a reliable channel for it to be stored safely in a server. Usually syslog messages
are transported using UDP protocol to the server which is not reliable. This calls for
the need for a reliable transport protocol like TCP to transfer the messages to the
syslog server.
IP SLA (Service-Level Agreement)
Cisco IP SLA enables continuous monitoring of network performance using
numerous network parameters, and it supports static routes, policy-based routing
(PBR), and dynamic routing protocols. Many types of probes may be configured to
monitor traffic in a network environment which include:

● Delay
● Jitter
● Packet loss
● Packet sequencing
● Path
● Connectivity
● Website and server download time
● Voice quality scores
The protocol determines whether the QoS on the network is sufficient to support IP
services