OSPF Operation and Route Selection - A detailed discussion

 OSPF Routing Protocol

 

Abstract:

              The report discusses the OSPF Routing protocol and its implementation in networks. There are different routing protocols supported by modern communication devices but how OSPF stands out as a unique protocol will be discussed in the report. OSPF   is a dynamic routing protocol developed by IETF International Engineering Task Force in 1989. OSPF uses a link-state algorithm known as the Dijkstra algorithm. The algorithm uses a mechanism to find the shortest path between the source and the destination. In link state routing the network topology and cost of all links are known. These link state packets are broadcasted to all the connected nodes in the network. OSPF divides the networks into different administrative domains called Areas. The reason behind this logical network division is to reduce the broadcasting in the whole network by limiting it to a specific group. OSPF grouping reasonably reduces the troubleshooting problems by confining each update or issue to the same area. In a bigger OSPF network, there can be multiple Areas these areas must be centrally connected to the Backbone or Area 0 network. This is Area 0 that helps in exchanging information between the other Areas and the concept is called Multi Area OSPF.

Introduction:

Routing is a layer-III concept in which information on routes is exchanged between different networks. The route information exchange between different networks is achieved by either adding the required manually or we use any protocol to do that. The concept of manual addition of routes to a router is called static routing and the routing exchange achieved by using a protocol is called dynamic routing.

OSPF is an interior gateway dynamic routing protocol that is used to exchange routing information between different networks but within the same Autonomous System. OSPF is one of the IP routing protocols that work on port number 89. For the initial database setup, it uses broadcasting on the IP address 224.0.0.5. The link state algorithm of OSPF collects link state information from all the connected networks which is called the link state database. OSPF advertises the directly connected networks for routing information exchange, after successful advertisement OSPF neighborships are established between the devices through which routing information’s exchanged.

OSPF is best suited for large complex networks that contain many sub-networks. These sub-networks are divided into Areas and all areas must be connected to Area0 for successful communication. Unlike RIP and other routing protocols, OSPF only shares the change that occurred in a network it doesn’t share any information with its neighbors when everything is running smoothly.

OSPF Operations

OSPF is a link-state routing protocol that is the most loved and famous enterprise network protocol. It is one of the best available Interior gateway protocols due to its unique features.

When OSPF is configured in a network for the first time, OSPF-enabled routers start listening to the neighbor OSPF-enabled routers for the collection of links state information. This link state information is gathered through the interfaces that connect one router to another OSPF router. At the end of the listening phase, the OSPF router saves the information to a database called link state database or LSDB. After gathering link state database information,

the OSPF router runs the Dijkstra algorithm or shortest path first (SPF) to convert the link state database to three different tables. One of the tables will be the one used for the shortest path calculation from the router to the destination network.

The OSPF operations can better be demonstrated using the below network diagram. Here we are having two networks connected to R1 and R5 and traffic has been generated from PC-A to PC-B. But before that OSPF routing protocol has been configured on the routers. All the routers in the below network belong to Area 0. Soon after the OSPF configuration, every router will try to find the neighboring router through the interface connecting to the neighboring router.

During this process, the OSPF routers will form neighboring adjacencies by going through different stages. Every router will form a neighbor adjacency with the other connected OSPF routers. After adjacencies, the router try to find the best path to PC-B from PC-A. OSPF cost of all the paths that connect it with Network PC- B. Based on this path cost OSPF router selects the best path towards the target network. The path having the lowest cost is considered the best path.

In the above scenario, the PC-A has 15+ paths and OSPF will select the best path by calculating the cost of each interface.

Based on the interface cost of interfaces, the router R1 will select the highlighted path of the PC-A traffic because the highlighted interface has the lowest cost. This is the best network route and the OSPF router will keep this routing information in the routing table. The routing will not change until and unless any change occurs in the network.

If any link between R1 and R5 doesn’t break path will remain the same but when a link failure occurs, OSPF will again run the Dijkstra algorithm to find the next best available path in the above diagram the values shown with each link is the link cost. The link cost is calculated with the formula below.

            cost = reference bandwidth / interface bandwidth

Here cost of the interface is calculated by dividing the reference bandwidth by the bandwidth of an OSPF-enabled interface. Below is the reference chart for OSPF costs.

Interface	Bandwidth	OSPF Cost
Gigabit Ethernet	1 Gbps	1
Fast Ethernet	100 Mbps	1
Ethernet	10 Mbps	10
E1	2 Mbps	48
T1	1.55 Mbps	64

 

Advantages of Configuring OSPF In a Network

·        There are no hop limitations in OSPF like we have in some other protocols like RIP and IGRP.

·        OSPF has a very fast convergence and calculates the next path in instants when the PR path goes down.

·        It divides the administrative networks into different areas for easy management and Maintenance.

·        Provides the effective use of VLSM.

·        Supports load balancing, Active and backup configurations.

OSPF Implementation and Configuration

OSPF protocol is used for large networks where network reliability, network Scalability, network redundancy is, and load balancing is valued. The implementation of OSPF in an enterprise network provides us with the above advantages.

When the OSPF is configured for networks that have a single group we call it single Area OSPF and when the network is a combination of many groups such a network is called Multi Area OSPF. In the below network we will configure a multi-area network OSPF network and analyze the routing tables.

 The above network of multiple paths when a packet travels from PC1 to PC2. The packet has three paths when it reaches R1. The paths have different costs due to different types of cables. The top path is having a longer path due to the number of routers in between but it is using fast ether cable which has a lower cost. The second path is a serial cable-based path which is the shortest path in the network. The last path is also a serial cable-based path. The routing table is then installed according to the shortest and best path. Each router has a database received from the connected router for the formation of routing table.

Each OSPF router is identified with a router ID which is an IP address and is the highest IP address of this router and being used for identification in the network.

For successful routing information, OSPF enabled router to establish neighborship adjacencies which can be queried as below. The neighbor ids are the routers with which this router has established neighborship relations.

All routers keep checking the neighborship adjacencies by sending hello and keep alive packets. They are set to a specific value and every OSPF router should follow these values for establishing adjacencies. If the hello, dead timer between two routers is not the same, the routers can’t exchange routing information between them. We can confirm the hello and dead time by querying the information of OSPF enabled router interface.

Now Let’s check which path the packet are choosing here when they arrive from the PC-A. The router routing table can be queried to check the status of router of the network 2.

We are receiving the required network 2 routers on R1 through the gigabit Ethernet0/0 which is connected to the top router which is the best path. If this path gets down the router R1 will select the next best path.

Authentication in OSPF

One of the best features that OSPF offers is the authentication that is done to secure the routing updates between OSPF routers. OSPF supports two types of authentication plain text authentication and encrypted MD5 authentication. For the proper function of authentication methodology, all routers should be enabled with the same authentication mechanism.

The Concept of OSPF Virtual Links

One of the mechanisms of OSPF usage is in the formation of Areas and connection of all Areas with the backbone Area or Area 0. But in the case were there is not possible to connect other areas with the backbone areas. In such cases, still, the other Areas are connected to area 0 through some workaround to build a logical connection with area 0.  This mechanism of virtual connectivity is via OSPF virtual links.

OSPF Table Types

OSPF maintains three types which are below.

·        Neighbour Table: Information of OSPF neighbors.

·        Topology Table: Information about the whole topology

·        Routing Table: List of the best route to different networks.

 

 

Conclusion

This report discusses the OSPF routing protocol, its implementation in a network, and the configuration steps with advantages. It also discusses the OSPF operations and convergence with the failure testing. We have developed a scenario where we have multiple paths from source to destination and the selection of the best path. We also tested when the best path gets unavailable and how OSPF selects the second-best path. The OSPF cost factor is one of the important factors that can be used for manipulation of the best path and checking the type of interfaces that have different costs and the formula used for cost calculation. On its advantages and other good features, it stands out as one of the best protocols and most used protocols for enterprise networks.