Читать книгу Understanding Infrastructure Edge Computing - Alex Marcham - Страница 54

To contextualise all of the routing information presented so far in this chapter, let us use an end‐to‐end example of traffic being routed across multiple networks from its source to its destination (see Figure 3.1). In this example, traffic will flow from a source on the left of the diagram to a destination on the right.

In this example, traffic will need to be routed across two networks which are peering using BGP as their EGP. One network is using OSPF for its IGP, while the other will use IS‐IS. The traffic is using IPv4, though the same concepts and flow seen in this example would apply to IPv6 traffic as well:

1 Traffic is generated by a device within the first network. Its destination is a server in the second network. The majority of internet use cases follow this general structure; examples include a device accessing a video or a cloud service from a remote server across networks.Figure 3.1 Routing process example.

2 The device sends its traffic to its local gateway, which is a router. This router inspects the IP packet headers of the traffic from the device and deduces that its neighbouring router is the best route available to reach this destination. The information to inform this deduction was the result of the router querying its local routing table, which was generated by OSPF using information received from the local router and its neighbouring routers across the network.

3 The traffic is sent to that neighbouring router, which then inspects its own routing table and determines that the best route to the destination of that traffic is via an external network. This same router has established a BGP peering with that external network, and so it sends the traffic to its BGP peer. Although it is an external network, this process is identical from a routing perspective to how traffic was moved within the network, excepting the use of BGP.

4 Once the traffic has entered the second network, the router which received it repeats the process of looking up the optimal route to the destination of the traffic in its routing table. This routing table informs the router that the optimal path is through its neighbouring router within the network, and so it sends the traffic over its link to that router. Note that although this network uses IS‐IS for its internal routing protocol, once the routing table is created by the routing protocol, the traffic routing process itself is identical on a hop‐by‐hop basis.

5 Finally, the last router receives the traffic and, upon inspecting its routing table, discovers that the subnetwork containing the destination of the traffic is directly attached to one of its local interfaces. The router sends the traffic over that link to its destination, and the process of routing traffic from source to destination is complete. Often traffic will be sent responding to the traffic which was sent; in this case, the same process is performed, from right to left.

Regardless of the size or number of networks between the source and destination of traffic, this is the routing process which takes place. Although different routing protocols differ in how they create their routing tables, how they pass information between each other to create them, and the criteria they use to determine the optimal route, this does not affect the general routing process as above.