- For selecting a path or route, a routing metric has
to be applied to a number of routes so as to select the best out of them.
- This
best route is called the optimal route with respect to the routing metric used.
- This routing metric is computed with the help of the routing algorithms in
computer networking.
- It consists of information such as network delay, hop
count, network delay, load, MTU, path cost, communication cost, reliability and
so on.
- Only the best or the optimal routes are stored in the routing tables
that reside in the memory of the routers.
- The other information is stored in
either the topological or the link state databases.
- There are many types of
routing protocol and each of them has a routing metric specific to it.
- Some
external heuristic is required to be used by the multi-protocol routers for
selecting between the routes determined using various routing protocols.
- For
example, the administrative distance is the value that is attributed to all the
routes in Cisco routers.
- Here, smaller distances mean that the protocol is a
reliable one.
- Host specific routes to a certain device can be set up by the
local network admin.
- This will offer more control over the usage of the network
along with better overall security and permission for testing.
- This advantage
comes handy especially when it is needed to debug the routing tables and the
connections.
In this article we discuss about the optimal routes.
- With the
growing popularity of the IP networks as the mission critical tools for
business, the need for methods and techniques using which the network’s routing
posture can be monitored is increasing.
- Many routing issues or even incorrect
routing can lead to undesirable effects on the network such as downtime,
flapping or performance degradation.
- Route analytic are the techniques and
tools that are used for monitoring the routing in a network.
The performance of
the network is measured using the following 2 factors:
- Throughput or the Quantity
of service: This includes the amount of data that is transmitted and time
it takes to transfer.
- Average packet delay or Quality of service: This includes the time taken by a packet to arrive at its destination and
the response of the system to the commands entered by the user.
- There is always a constant battle
between the fairness and optimality or we can say between quantity of service
and quality of service.
- For optimizing the throughput, the paths existing
between the nodes have to be saturated and the response time from source point
to destination point must be noticed.
For finding the optimal routes, we have
two types of algorithms namely:
- Adaptive Algorithms: These algorithms are
meant for the networks in which the routes change in a dynamic manner.
Here the information regarding the route to be followed is obtained at the
run time itself from adjacent as well as the all other routers. The routes
change whenever there is a change in the load, change in the topology and
every delta T seconds.
- Non – adaptive algorithms: These algorithms
the same routes cannot be followed every time. Therefore the measurements
that were made for the previous condition cannot be used for the current
condition. The routes thus obtained are called static routes and are
computed at the boot time.
Finding optimal routes requires
following the principle of optimality according to which the optimal path
between an intermediate router and the destination router lies on the same
route from the source to the destination route.
