Compare and Contrast Link State and Distance Vector Protocols.
Link State Protocols: Link state
protocols are those protocols in which every router knows the topology of the
network and the cost of every link. This information is transmitted to one
router from its neighboring routers via broadcasting the link state packets.
Link state packet contains the information of link cost and identities of every
directly connected network.
After gathering all the network
information every router executes an algorithm called the Dijkstra Algorithm to
find the shortest path to every network considering the link cost from the source
to the destination network.
Distance Vector Protocols: These are
the protocols that receive information from its directly connected neighboring
routers only. Thus, they have a local view they do not know about the network
topology but rather depend on the information received from their neighboring
routers. The distance vector routers share their whole routing table
periodically every 30 mins. When a router receives different information
from the previous information it recalculates its routing table again.
Advantages and Disadvantages of
Link state and Distance Vector Protocols
Link State Routing Protocols
|
Link State Protocols |
|
|
Advantages |
Dis-Advantages |
|
Best suited for large networks due to good scalability |
Link-state has smaller routing tables |
|
and lower convergence time makes without routing loops due to the availability
of a global view. |
Do not share updates until any change in the network |
|
Link state databases can be minimized by diving the network into
different areas. |
for a fluctuating network, updates are shared continuously which
even chokes the network |
|
Link state packets are well sequenced and show the age of the last
update also. |
Needs good network knowledge to implement link state routing
protocols. |
Distance Vector Protocols
|
Distance Vector Protocols |
|
|
Advantages |
Dis-Advantages |
|
Less expertise is required for implementation |
Slow convergence time, routing updates are shared after every 30
s. |
|
Less resource consumption like low CPU and memory is required |
Not suited for large networks due to slow convergence and limit
to 15 hops only |
|
Suited for small networks as less bandwidth is required to send
routing updates as the size of packets is small |
Routing loops can occur due to slow convergence during a network
change. |
Routing Loop
The routing loop is one of the
major issues that may occur in a network. In a routing loop, a packet is
continuously routed through the network without delivering it to a destination.
Thus, the packet routes in a circle endlessly and can prove damaging for the network
by consuming network bandwidth unnecessarily.
If a packet enters a network with
loops and loops are already consuming a good part of the bandwidth, then the
new packet to a different network will also face delays and high latency. If
the packet belongs to the already looped network, then the packet will route in
a loop contributing to destroying the network performance.
We have a few methods to avoid
routing loops, as mentioned below.
·
Maximum Hop Count
·
Spilt Horizon
·
Route poisoning
·
Hold down timers.
These technologies are already
existing in the protocols are we can fine-tune a few parameters like
manipulating the hold-down timer etc. to avoid routing loops.
Routing Protocol Path Selection.
Every protocol has a different metric for the best path selection and a metric is a value that every protocol assigns to every path depending upon its usefulness.
The routing protocol RIP has a metric of Hop count which means that it does not check the quality of a link, it just checks the least number of routers between the source and destination networks.
In the case of OSPF routing protocol, it must do nothing with a number of hops, it checks the bandwidth and delays on a link, or called the cost of a link. It selects the path with the least cost.
Therefore, the metric of protocols is not comparable and under the same conditions, they select different paths.
No comments:
Post a Comment