Networking Heart: IP Routing

IP Routing:

Manual configuration
Routes must be configured both the ways .ie. for both sending data and receiving data
Static has a lower administrative distance of 1. Thus, static route will be preferred.
If any router fails, then, there will be no path. Thus, static fails
Also, congestion will happen on a specific route if lot of traffic is being sent. Still, static route is taken
In Production, static routes are used as backup

Automatic selection of routes using algorithms
Algorithms may be dynamic, where the routers make decisions based on information they gather, and the routes change over time, adaptively.
Default route means we just send all traffic to the default router which becomes the next hop. Using default route is useful since it saves memory (instead of giving a next router for all routes) and also it helps to send packets to unknown router using the default router

--------------------------------------------------------------------------------------------------------------------------------

IP Addressing

To find the device: we need identification and location. The identification is the host id. The location is the network id.
If we have more users, we need more host bits.

Class A- 8 n/w bits and 24 host bits (0-126) (127 is used for loopback, so can’t be used for hosts)
Class B- 16 n/w bits and 16 host bits (128-191)
Class C- 24 n/w bits and 8 host bits (192-223)
Class D-multicast (224-239)
Class E- experimental (240-255)^[a]

Class A: We have 8 n/w bits. But, the first bit is fixed as 0. Thus, we have only (2^7) networks possible. No of hosts= (224)-2 .
Class B: We have 16 n/w bits. But, the first bit is fixed as 10. Thus, we have only (2^14) networks possible. No of hosts= (216)-2
Class C: We have 24 n/w bits. But, the first bit is fixed as 110. Thus, we have only (2^21) networks possible. No of hosts= (28)-2^[b]

Note: the (-2) in no. of hosts is because one address is for n/w address and another one is for broadcast address

Subnet Mask is used to tell the number of bits for network part.
Eg) 10.0.0.0?- is class A. Therefore, subnet mask is 255.0.0.0. The first host address is 10.0.0.1 and the last host address is 10.255.255.254.
Eg) 172.16.0.0 ?- Class B therefore subnet mask is 255.255.0.0 . First host address is 172.16.0.1 and last is 172.16.255.254
Eg) 192.168.1.0 ?-is class C. hence, subnet mask is /24. First host ID is 192.168.1.1 and last is 192.168.1.254. The broadcast address is 192.168.1.255
Each Class has some range for private IP address:

Zero Subnet:It is the first subnet obtained when subnetting the network address. Eg) 192.168.1.0/24 is the zero subnet for class C. Nowadays, we can assign this IP address to hosts also since we use classless IP instead of classful IP addressing.
Public IP address must be unique and is assigned by the ISP

--------------------------------------------------------------------------------------------------------------------------------

Subnetting:

If we use only classes, then, either we have a lot of hosts or a lot of networks. Thus, IP addresses are wasted. Hence, we use variable length subnets.
In subnetting, we borrow the bits from the host part.
Eg) 10.0.0.0/8.

Here, the n/w part is 10.
Host part is 0.0.0. We borrow bits from here. Let us take 2 bits. SO, host part starts with (XX000000.)b
Thus, we get 4 subnets 10.0.0.0/10, 10.64.0.0/10, 10.128.0.0/10, 10.192.0.0/10

Hint: The subnets will be in multiples of the first subnet. Eg) 64*1=61, 64*2=128, 64*3=192
If we use /31, then, we will have only one host in the subnet (if we consider broadcast address) or 2 hosts if we don’t consider broadcast. Since a point to point (between 2 routers on end of the n/w) doesn't really need a broadcast address, /31 can be used here.
CIDR- using only subnets instead of classes

Thursday, 28 June 2018