4.2 Steps to OSPF Operation
4.2.2 Step 1: Establishing router adjacencies
Because OSPF routing is dependent on the status of a link between two routers, neighbor routers must "recognize" each other on the network before they can share information. Neighbor routers are routers that share a common data link, as shown in the figure. This process is done using the Hello protocol (part of the OSPF suite of protocols). This protocol enables neighbor routers to establish virtual point-to-point links called adjacencies and to ensure bi-directional communication between neighbors before exchanging link-state information.
Lab Activity
  In this lab, you will learn how to enable OSPF routing protocol in Area 0 only.

Hello packets are sent periodically out of each interface using IP multicast addresses. Multicast addresses are used to send packets to a group of devices, rather than a single router (unicast) or the entire network (broadcast). The information contained in a Hello packet is as follows:

  • Router ID --- Used to identify a router to OSPF. The highest IP address on an active physical or loopback interface is used as the Router ID. This identification is important in establishing neighbor relationships and coordinating messages between OSPF routers on the network. Also, the Router ID is used to break ties during the DR and BDR election processes if the priority values are equal. (DR and BDR are discussed later.)
  • Hello interval --- Specifies the frequency in seconds that a router sends out Hello messages (the default is 10 seconds on multiaccess networks), as shown in the Figure. The Hello interval must be the same on neighboring routers.
  • Dead interval --- The dead interval is the time in seconds that a router waits to hear from a neighboring router before declaring the neighbor down (four times the Hello interval by default). The Dead interval must be the same on neighboring routers.
  • This Hello process also provides quicker detection of failed routers because Hellos are exchanged every 10 seconds. Routers expect to hear from their neighbors every 10 seconds. If a router is silent for 40 seconds, its neighbors believe it is down. Note: This is true if you assume default configuration values.
  • Neighbors --- The neighbors to which adjacencies have been created. (At the initial startup point, this field is empty because the first router has not received Hello messages from other routers.)
  • Area-ID --- The common identification number for an area. To communicate, two routers must share a common network segment, have interfaces configured to belong to the same area on that segment, and share the same subnet address and mask. Routers from the same area will each have the same link-state database information.
  • Router Priority --- A value that indicates the priority of this router when selecting a Designated Router (DR) and Backup Designated Router (BDR). The default priority is 1 and can be configured to a higher number to ensure a specified router becomes the DR.
  • DR and BDR --- The IP addresses of the DR and BDR for the specific network, if known.
  • Authentication password --- If authentication is enabled, two routers must exchange the same password. Authentication is optional.
  • Stub area flag --- Indicates the presence of a stub area. A stub area is a special area that has only one router that connects to another area. Stub areas are discussed in the next chapter. Two routers must agree on the stub area flag in the Hello packets.

Neighbor relationships differ based on the types of connection that exist between the routers , as follows:

  • Broadcast multiaccess --- In a typical broadcast LAN environment such as Ethernet or Token Ring, the OSPF routers communicate with the DR by using multicast address 224.0.0.6 (aka All DR Routers).. The DR sends this learned information back to the OSPF routers by using multicast address 224.0.0.5 (aka All SPF Routers). Neighbor relationships are formed based on a dynamic learning process using multicast Hellos (explained earlier).
  • Point-to-point --- Across a serial connection, for example, neighbor relationships are formed only between two routers on each point-to-point link. Because there are only two routers on this type of connection, no DR or BDR is needed or used.
  • Nonbroadcast multiaccess (NBMA) --- Across a standard Frame Relay or ATM network, there is no broadcast capability. As such, all OSPF messages in a nonbroadcast environment are sent as unicast. As such, neighbor relationships often require additional configuration. However, OSPF routers on an NBMA network do elect a DR and a BDR.

The exchange process, when all routers come up on the network at the same time, uses the Hello protocol. This process is detailed in the following steps:

  1. Router A is enabled on the LAN, but OSPF is in the down state because it has not exchanged information with any other router. Router A sends a Hello packet containing information about itself to the network. This information includes Router A's Router ID.

  2. All OSPF routers with the IP multicast address enabled receive the Hello packet from Router A and store the information about Router A locally. This is the Init state.

  3. All routers that received the Hello packet send a unicast reply Hello packet to Router A with their corresponding information. A unicast packet is a packet addressed to one specific device on the network. The information that the responding routers send includes details about all neighboring routers, including Router A.

  4. When Router A receives these reply packets, it adds all the routers that had its (Router A's) Router ID in their packet to its own database. This step is referred to as the two-way state. At this point, all routers that have learned about each other establish bi-directional communication.

  5. Next, neighbor routers select the DR and BDR. The DR and BDR election process is described in the "Step 2" next section. This process must occur before routers can begin exchanging the complete link-state information (information about the network topology and routing devices). Link-state exchanges are discussed in the "Step 3" section to follow.

  6. Neighboring routers then determine the master/slave relationship between themselves (e.g. who will initiate sending the link-state advertisements?), exchange LSAs that describe each router's link state database, and establish full adjacency.

  7. Periodically (10 seconds by default), the routers within a network exchange Hello packets to ensure communication is still working. These Hello "keepalive" updates include the DR router ID and the list of routers whose Hello packets have been received by the router sending the Hello.