Understand the Routing Process, Part 3
View PDF | Print View
by: Guest
Total views: 12
Word Count: 593
Routers route, that\'s what they do. A packet comes in and a packet goes out. What happens to the packet containing your data inside the router? How does the router decide which output interface to transmit the packet out of? What if the router has two routes to the same destination how does it deal with that decision making process?
adult shopping | vibrators online | clitoral vibrators
Dynamic Routes
Static routes are fine for small networks or configuring backup links. Once routed networks start growing using static routes to maintain a stable network is not scalable and becomes very difficult to maintain. So we use dynamic routing protocols.
Routing protocols come is various shapes and sizes some are very simple to understand, like two stoke engines i.e few moving parts, make a lot of noise and don't get you very far and some are very sophisticated, very fast, route across huge networks but can also be very complicated to fault find if not configured correctly.
Dynamic routes work by having routers exchange network information. Initially routers can only advertise networks which are directly connected, once a router receives an advertised network it can then if configured to do so will in turn advertise it to it's directly connected neighbours setting it's self as the next hop to get to the destination network.
When a router is offered routes it will place them into the routing table with a label indicating how the route was learnt. For example if the route is a directly connected interface the label against the route is "C" denoting a "Connected" interface, "S=Static", "D=EIGRP", "O=OSPF", "R-RIP".
The same process a router goes through to decide which static routes to choose between in the event that a router has more than one route to the same destination, i.e the longest prefix match, the router will use the same process if the route was learnt dynamically or statically.
If a router has two routes to the same destination it will choose the route with the closest match. If one of the routes is static and another is dynamic; what then?, we would like the router to choose the dynamic route, but the Administrative Distance of the static route is by default set to 1, dynamic routing protocols have higher Administrative Distances, here we will have to increase the value of the Administrative distance of the static route or remove the route altogether. Administrative Distances of most popular dynamic routing protocols are as follows
RIP=120
OSPF=110
EIGRP=90
Matching Prefix lengths, Matching Administrative Distances, now what?
The router is presented with two or more routes with the same prefix length and from the same routing protocol, how does the router then choose between all of these routes?.
Every routing protocol determines the desirability of a path by using a metric, different routing protocols use different metrics, some use simple metrics, RIP uses simple HOP COUNT, one router equals one hop, OSPF uses COST, cost is the inverse of the configured interface bandwidth. EIGRP uses a composite of interface BANDWIDTH, DELAY, RELIABLITY, LOAD and MTU, (Although only bandwidth and Delay are used by default).
When a router has to choose between two or more routes to the same destination it will choose the route with the lowest metric.
If the metrics are the same, the router will then share the traffic across the paths either equally or unbalanced which means traffic is shared across the paths dependant on the ratio of the metrics.
To summarise a router will make routing decisions as follows:
1. Longest match irrelevant of how the route was learnt
2. Lowest Administrative distance
3. Lowest Metric
4. Load balance
About the Author
Rating: Not yet rated