What Routers actually do in routing?
- Routing can be mainly divided into two pieces.
- They are direct routing and indirect routing.
· Using direct routing we can send messages to a machine in the same LAN.
· Using indirect routing we can send messages between LANs.
If a data packet travels through a number of routers to reach the destination. In this routing method, all routers should maintain the routing table and they should use routing strategies to maintain the data flow.
Four main routing strategies
- fixed routing
- random routing
- adaptive routing
In the fixed routing, routing data are stored in a central location.The main advantage in this strategy is new information can be easily updated by directly access the central location. But there are some disadvantages in this method.
- Each and every IP packet should be analyzed
- Network traffic may increase at the central database because it has to serve lots of requests from routers
- Dynamic changes are not possible
Second method is flooding. In this method information will be send to all paths of the network. This method causes high possibility of network data traffic and duplicate data might be accepted by the destination. But also most likely data will reach the destination.
In random routing method data routers randomly selects a path to send and send the data packet on that path directly. This method will decrease the data traffic. But there is no guarantee about the data packet will reach the destination or not.
In adaptive routing, each router maintains routing table and it changes automatically when network receives data updates. With this method, network traffic will be minimized and the best route can be verified easily. Speed of the data flow will be increased.
Routing methods used in adaptive routing
- Next hop routing
- Host specific
- Network specific
2. Default routing
- Next hop routing -> Host specific
* Each router keeps a table entry for each host
* Each table entry has Host IP and the Interface
• Large number of records (if multiple paths are available, number of records will be increased)
• Table updating is difficult as it should be done for each and every host.(if the host IP changes)
- Next hop routing -> Network specific
* Each router keeps a table entry for each network (one record for one network)
* Table entry has Network address and Interface
• Number of records are limited (Table updates are not for each host, but for a network)
• Update is easy
- Default Routing
This is just another record in the routing table to say if any of the records does not match with the IP packet destination IP what path should be taken
You can only use default routing on stub networks, which means that they have only one exit port out of the network
Router(config)#ip route [<next hop address> <space><subnet mask of next hop><space> <exit interface>]
Routing Table update Methods
Basically there are three. They are connected, static and dynamic.
1. Once the router is connected to the network its interfaces are given IP addresses.
2. With that router automatically identifies the network addresses of networks which its connected.
1. Administrator can manually give the routing table records
Router(config)#ip route [<destination network><destination network subnet mask> <next hop address | exit interface>
o Minimal CPU processing,
o Easier for administrator to understand and configure
o Configuration and maintenance is time-consuming and error-prone.
o Administrator inspection is required to maintain changing route information
o Does suitable for growing networks because maintenance becomes a hard job
o Requires complete knowledge of the whole network for any implementation
1. By using some protocols routing tables can be updated automatically
2. Initially we might have only ‘connected’ records then we might add few ‘static’ records then it will get dynamic updates automatically.
o Administrator has less work maintaining the configuration when adding or deleting networks
o Protocols automatically react to the topology changes and less error-prone
o More scalable, growing the network usually does not present a problem
o Router resources are used (CPU cycles, memory and link bandwidth)
o More administrator knowledge is required for configuration, verification, and troubleshooting
We use Routing Protocols to
ü Immediately update the routing tables of all routers in any change of the network like an addition or a removal or in a fault.
ü If there are many routes to a destination, the best route should be selected or share the traffic through a different route.
Dynamic Routing Protocols