EIGRP
Enhanced Interior gateway protocol
It is also known as Hybird, because it possesses the characteristics of both Distance vector (RIP, IGRP) and Link state (OSPF,BGP,IS-IS).
It is a classless chanced distance vector protocol (supports sub netting). Means, distance vector is a class full protocol which does not support subletting. Whereas, EIGRP support sub netting. It uses the autonomous system to describe routes. An autonomous system is assigned a globally unique number. Also called (ASN). Networks within the same autonomous system share routing information with each other.
It send updates if there is any change. Other protocols like rip has a default time of 30 secs. After this time Rip will send its complete routing table with updates. Consider that for a large network. Rip protocol will consume all the resources.
It uses a subnet mask in its router updates, a characteristic of linkstate (VLSM).
What is VLSM:
It stands for variable length subnet mask. It removes the restriction for any class. You can chose any class and subnet it to acquire your desired number of host and networks.
E.g For a class A ip
10.65.23.100 in actual it will come up with /8. The subnet mask will look like 255.0.0.0. in this case you have first 8 bits (1st octet )fixed. And other 24 bits (3 octets) are available to be used. Just think now how many ips you might waste if you use the same traditional method with VLSM. For class A you use 2 ^16 number of IPS.
Even for a big organization this number is too large to fit. VLSM is introduced to save IPS. With any class you can acquire the number of ips you need and save the rest of them.
Let’s take the same example of Class A
Actual IP Range: 10.65.23.100/8
After sub netting:
IP: 10.65.23.100/28
In this case you are using class A ip. But the number of use able IPwill be 16 -2 =14, 1st subnet id, 2nd broadcast ip, 1st and last ips we cannot use.
10.65.23.100/28
The subnet mask will be 255.255.255.240
Number of networks: 2^4=16
Number of host: 2^4=16-2=14
Therefore by using the same Class A ip we only uses 16 ips. Otherwise with class A there comes 2^24 ips.
EIGRP uses a diffusing update algorithm (DUAL), to find the best path. Among all feasible paths. It also ensures a loop free path.
EIGRP traffic sends as unicast or multicast on address 244.0.0.10 depending upon the packets.
It uses the same autonomous system to from neighbor relationships.
It uses RTP(reliable transport protocol) to ensure the delivery of packets.
It will not send periodic updates.if there is any change then it will send the update only for that change.
It supports, IP, IPX and Apple talk
The administrative distance of EIGRP IS 90.
It uses the same metric like IGRP. (bandwidth, delay, reliability, load, MTU). Bandwidth and delay are used to calculate its distance. The others reliability load and MTU can also be used.
It keeps backup routes:
EIGRP keeps a back path along with the primary path. Incase, the primary get down it will automatically shifts to back up route. Where as OSPF and IS-IS do not store backup paths. If the primary get down they will run the algorithm on all routes stored in their table to find the alternative path.
Unequal cost load balancing:
Load cost balancing means if there is extra load on one route (primary route) then you can share some of the packets with other routes e.g (backuproute) to deliver the packets to the same destination.
Unequal means you will decide how much data you like to transmit through one route and how much will got through other route. In the end both will reach to the same destination.
In other protocols, they are using the equal load balancing. Means both routes will share equally. E.g if there are four packets then 2 will go through primary and two will go through backup. Here in EIGRP, you can distributes the packets as per your requirements . e.g 3 packets from primary route and 1 packet from backup route.
The load balancing is based on the bandwidth. It means the division of the bandwidth among different routes.
Example 1:
Load cost balancing means if there is extra load on one route (primary route) then you can share some of the packets with other routes e.g (backuproute) to deliver the packets to the same destination.
Unequal means you will decide how much data you like to transmit through one route and how much will got through other route. In the end both will reach to the same destination.
In other protocols, they are using the equal load balancing. Means both routes will share equally. E.g if there are four packets then 2 will go through primary and two will go through backup. Here in EIGRP, you can distributes the packets as per your requirements . e.g 3 packets from primary route and 1 packet from backup route.
The load balancing is based on the bandwidth. It means the division of the bandwidth among different routes.
Example 1:
Example 2:
EIGRP Tables and Terminology:
EIGRP has three tables
1) Neighbor table
2) Topology table
3) Routing table
Neighbor table:
Each router will keep the records of its neighbors. Hello packets are used to introduce and from relationship
Hello packet is sent as multicast on 244.0.0.10
The neighbor relationships are also known as agjencies . After adjencies are formed the routers can share information’s.
On a high-speed LAN or WAN the hello packets are send after every 5 sec’s by default.
On slow speed wan (T1 or slower) the hello packets are send after every 60 sec’s by default.
Configuration (To change the speed of hello packets)
R1(config)#ip hello-interval eigrp 10 7
Where 10 is the autonomous number and 5 seconds are changed to 7 seconds.
Hold timers:
This is used to indicate how long a router should wait before deciding the neighbor is inactive, if it is not receiving the hello packets.
By default it is three times the hello packet.
For high speed it is 15 sec
For slow speed it is 180 sec
Configuration (To change the hold timer)
R1(config)#ip hold-interval eigrp 10 21
Where 10 is the autonomous number and 15 seconds are changed to21 seconds.
The neighbor table is constructed on hello packets. Hello packets consists of the followings
1) The IP address of the neighbor route.
2) The hold timer.
3) The local interface through which the hello packets are received.
4) The autonomous number in which the neighbors are present.
Configuration (Log messages and errors)
R1(Config)#router eigrp 10
R1(Config-router)#eigrp log-neighbor-changes
R1(Config-router)#eigrp log-neighbor-warnings
Topology Table:
The topology table is based on the followings:
1) The feasible distance
2) The successor
3) The advertise distance
4) The feasible successor.
Note that the above mentioned are placed in the order they will appear, one after another , in reality . After that there are two more.
5) Active route
6) Passive route
The feasible Distance:
Every route to destination has its feasible distance. It is the metric used by EIGRP to select the best path. If there are multiple ways to reach the same destination then path with the lowest feasible distance will be selected and it becomes the successor.
Successor:
It will be the best distance selected among different routes to from source to destination e.g from A to H in below example.
EIGRP Tables and Terminology:
EIGRP has three tables
1) Neighbor table
2) Topology table
3) Routing table
Neighbor table:
Each router will keep the records of its neighbors. Hello packets are used to introduce and from relationship
Hello packet is sent as multicast on 244.0.0.10
The neighbor relationships are also known as agjencies . After adjencies are formed the routers can share information’s.
On a high-speed LAN or WAN the hello packets are send after every 5 sec’s by default.
On slow speed wan (T1 or slower) the hello packets are send after every 60 sec’s by default.
Configuration (To change the speed of hello packets)
R1(config)#ip hello-interval eigrp 10 7
Where 10 is the autonomous number and 5 seconds are changed to 7 seconds.
Hold timers:
This is used to indicate how long a router should wait before deciding the neighbor is inactive, if it is not receiving the hello packets.
By default it is three times the hello packet.
For high speed it is 15 sec
For slow speed it is 180 sec
Configuration (To change the hold timer)
R1(config)#ip hold-interval eigrp 10 21
Where 10 is the autonomous number and 15 seconds are changed to21 seconds.
The neighbor table is constructed on hello packets. Hello packets consists of the followings
1) The IP address of the neighbor route.
2) The hold timer.
3) The local interface through which the hello packets are received.
4) The autonomous number in which the neighbors are present.
Configuration (Log messages and errors)
R1(Config)#router eigrp 10
R1(Config-router)#eigrp log-neighbor-changes
R1(Config-router)#eigrp log-neighbor-warnings
Topology Table:
The topology table is based on the followings:
1) The feasible distance
2) The successor
3) The advertise distance
4) The feasible successor.
Note that the above mentioned are placed in the order they will appear, one after another , in reality . After that there are two more.
5) Active route
6) Passive route
The feasible Distance:
Every route to destination has its feasible distance. It is the metric used by EIGRP to select the best path. If there are multiple ways to reach the same destination then path with the lowest feasible distance will be selected and it becomes the successor.
Successor:
It will be the best distance selected among different routes to from source to destination e.g from A to H in below example.
Let take the above example.
We have multiple routes from A to reach H. the question is this how we select the best route from A to H.
Here comes the feasible distance to solve this problem. As we note each route has its own cost e.g from A to B = 8 , A to C= 4 and so on. Each route has its own cost called the feasible distance from A to H.
To find FD
Feasible distance from A to H of routes : A+B+E+H=16
Feasible distance from A to H of routes : A+C+F+H=27
Feasible distance from A to H of routes : A+D+G+H=11
As we can see the lowest figure is 11, this will be the feasible distance from A to H. This F.D is less than the other F.D’s therefore this will also be selected as the successor .
Successor:
As from the above we find out that among the multiple paths the lowest feasible distance is the successor.
Feasible successor:
Now its’ time to find out the backup route also known as feasible successor:
To find out F.S we depends on advertise distance (A.D)
Advertise distance:
Advertise distance is the distance from the neighbor route up to the destination.
Like
Advertise Distance from: B to H= 2+6=8
Advertise Distance from: C to H=14+9=23
Advertise Distance from: D to H=5+4=9
Here the lowest one is B to H=8, that is also less than the F.D therefore it is selected as the backup route known as Feasible successor in topology table. In case the successor goes down the feasible successor will takes it place as a backup.
Note here again to be selected as a feasible successor the advertise distance should be less than the feasible distance. This is to prevent loops in the network.
Active Route:
This means the router is searching for its backup. During this period it will put A in its table and this is also bad.
Passive route:
This means everything is fine and you will see P.
EIGRP PACKETS:
Packets are used while visiting the neighbors.
The types of EIGRP packets.
Hello packet
Update
Query
Reply
Acknowledgement.
1) Hello Packet:
Hello packets are used to form neighbor relationship. Let’s take real world example, before starting conversation with other person you need formal way to start with hi, hello. Same here is the case, hello packets are used to get to know each other at the beginning . later on they are used to keep in touch. To know the other end is up and working. As we mentioned above for high speed , time interval of hello packet is 5 sec and for slow speed it is 60 secs. They are sent on multicast link 244.0.0.10
2) Update Packets:
After the formal introduction done by hello packets and forming neighbor relationships, update packets will come to play their role. At first, the routing table will exchanged through update packets. One router will send all its routing table to other router by means of update packets. And in reply it will also receive the routing table of other routers by means of update packets. Later on, after exchanging the routing table, the update packets will only be used if there is any change. And at that time they will not start sending again the whole routing table (unlike rip). They will only send the specific update. Otherwise send only hello packets to keep in touch. The update is sent on multicast link 244.0.0.10
3) Query Packets:
If the successor is down and there is no alternative feasible successor available. The router will send a query to its neighbor to ask whether he is having any a route to reach to the network. This is sent on multicast link 244.0.0.10
4) Reply Packets:
In reply to query the router will send a query packet to let the first router know if they have a valid router available or no. this is sent on unicast link.
5) Acknowledgement Packet:
They are used for the reliable delivery of data. It has no data but only the acknowledgement. They are used for update, query and reply. The update, query and reply uses RTP (reliable transport protocol) . Means after delivery of packets they want to be sure that the data is received by the other end. In reply the other router will send the acknowledgement that ok I received what you sent to me.
EIGRP STATUS:
Active
Passive
Stuck-in-active(SIA)
Active:
Active means if the successor and any feasible successor fails or they are still looking for successor or feasible successor they will placed in active and you will find “A” that means it is still searching and in active state.
Passive:
If the network is reachable and they are fully converged they will placed in passive state. You will find “P” that means passive
Stuck-in-active(SIA):
Mostly the above two are the common states, but there is another state know as Stuck-in-active(SIA). If any router sends any query and does not receive reply in 3 minutes then it is in Stuck-in-active(SIA)
View current state of router in EIGRP:
Router# show ip eigrp topology
View only active routes:
Router# show ip eigrp topology active
We have multiple routes from A to reach H. the question is this how we select the best route from A to H.
Here comes the feasible distance to solve this problem. As we note each route has its own cost e.g from A to B = 8 , A to C= 4 and so on. Each route has its own cost called the feasible distance from A to H.
To find FD
Feasible distance from A to H of routes : A+B+E+H=16
Feasible distance from A to H of routes : A+C+F+H=27
Feasible distance from A to H of routes : A+D+G+H=11
As we can see the lowest figure is 11, this will be the feasible distance from A to H. This F.D is less than the other F.D’s therefore this will also be selected as the successor .
Successor:
As from the above we find out that among the multiple paths the lowest feasible distance is the successor.
Feasible successor:
Now its’ time to find out the backup route also known as feasible successor:
To find out F.S we depends on advertise distance (A.D)
Advertise distance:
Advertise distance is the distance from the neighbor route up to the destination.
Like
Advertise Distance from: B to H= 2+6=8
Advertise Distance from: C to H=14+9=23
Advertise Distance from: D to H=5+4=9
Here the lowest one is B to H=8, that is also less than the F.D therefore it is selected as the backup route known as Feasible successor in topology table. In case the successor goes down the feasible successor will takes it place as a backup.
Note here again to be selected as a feasible successor the advertise distance should be less than the feasible distance. This is to prevent loops in the network.
Active Route:
This means the router is searching for its backup. During this period it will put A in its table and this is also bad.
Passive route:
This means everything is fine and you will see P.
EIGRP PACKETS:
Packets are used while visiting the neighbors.
The types of EIGRP packets.
Hello packet
Update
Query
Reply
Acknowledgement.
1) Hello Packet:
Hello packets are used to form neighbor relationship. Let’s take real world example, before starting conversation with other person you need formal way to start with hi, hello. Same here is the case, hello packets are used to get to know each other at the beginning . later on they are used to keep in touch. To know the other end is up and working. As we mentioned above for high speed , time interval of hello packet is 5 sec and for slow speed it is 60 secs. They are sent on multicast link 244.0.0.10
2) Update Packets:
After the formal introduction done by hello packets and forming neighbor relationships, update packets will come to play their role. At first, the routing table will exchanged through update packets. One router will send all its routing table to other router by means of update packets. And in reply it will also receive the routing table of other routers by means of update packets. Later on, after exchanging the routing table, the update packets will only be used if there is any change. And at that time they will not start sending again the whole routing table (unlike rip). They will only send the specific update. Otherwise send only hello packets to keep in touch. The update is sent on multicast link 244.0.0.10
3) Query Packets:
If the successor is down and there is no alternative feasible successor available. The router will send a query to its neighbor to ask whether he is having any a route to reach to the network. This is sent on multicast link 244.0.0.10
4) Reply Packets:
In reply to query the router will send a query packet to let the first router know if they have a valid router available or no. this is sent on unicast link.
5) Acknowledgement Packet:
They are used for the reliable delivery of data. It has no data but only the acknowledgement. They are used for update, query and reply. The update, query and reply uses RTP (reliable transport protocol) . Means after delivery of packets they want to be sure that the data is received by the other end. In reply the other router will send the acknowledgement that ok I received what you sent to me.
EIGRP STATUS:
Active
Passive
Stuck-in-active(SIA)
Active:
Active means if the successor and any feasible successor fails or they are still looking for successor or feasible successor they will placed in active and you will find “A” that means it is still searching and in active state.
Passive:
If the network is reachable and they are fully converged they will placed in passive state. You will find “P” that means passive
Stuck-in-active(SIA):
Mostly the above two are the common states, but there is another state know as Stuck-in-active(SIA). If any router sends any query and does not receive reply in 3 minutes then it is in Stuck-in-active(SIA)
View current state of router in EIGRP:
Router# show ip eigrp topology
View only active routes:
Router# show ip eigrp topology active