NS tech Notes

NightShade1

Here i will post some of my mini tutorial or notes i have posted here and in other forums
Hope it helps ppl hehe

Router summarization

Ripv1 and IGRP
those protocols do auto summary to the classfull network which means
if you try to add a network of 10.1.1.1 it will auto summary to 10.0.0.0

You cant turn off this auto summary in Ripv1 and IGRP

you can turn it off on Ripv2 and EIGRP summarizing

OSPF and IS IS does not auto summery

Note:
Classful routing protocols (RIPv1 and IGRP) automatically summarize routes on the classful network boundary and do not support summarization on any other bit boundaries. Classless routing protocols support summarization on any bit boundary.


Route Summarization Calculation example

suppose you got this networks attached to your router like different LAns

172.16.12.0/24

172.16.13.0/24

172.16.14.0/24

172.16.15.0/24

1) first step:
You will convert everything to binary

10101100.00010000.00001100.00000000 -->172.16.12.0
10101100.00010000.00001101.00000000 --->172.16.13.0
10101100.00010000.00001110.00000000 -->172.16.14.0
10101100.00010000.00001111.00000000 -->172.16.15.0


2) Second Step

You need to locate the bits till where the bits are equals
What i mean with this ?

10101100.00010000.00001100.00000000 -->172.16.12.0
10101100.00010000.00001101.00000000 --->172.16.13.0
10101100.00010000.00001110.00000000 -->172.16.14.0
10101100.00010000.00001111.00000000 -->172.16.15.0

As you can see all the bits are equal till 6th bit of the 3rd octet(can you can see it above i BOLDED all the equals bits)

Now you find this we will go to step 3

3)Third step

You need to count those bits, the summary route number will be the first IP address in the block, and the subnet mask will be the number of common bits.

so the answer is
172.16.12.0/22


Well i think this is all you need to know about summarizing If anyone find any mistake on my explanation they are welcome but im pretty sure its like this.

NightShade, All right reserved

NightShade1

How does RIP fill their routing tables ???

ill give you a fast example

R1

---

R2

---

R3

---

R4

R1-R2 link has the network 10.1.1.0 network
R2-R3 link has the network 10.1.2.0 network
R3-R4 link has the network 10.1.3.0 network

In the beggining before sending the broadcast of their full routing table they will be like this
Router 1 table
10.1.1.0 directly connected

Router 2
10.1.1.0 directly connectd
10.1.2.0 directly connected

Router 3

10.1.2.0 directly connected
10.1.3.0 directly connected

Router 4
10.1.3.0 directly connected

They start exchanging routing updates so they will start filling up their routing tables

Router 1
10.1.1.0 directly connected
10.1.2.0 via 10.1.1.1 hop 1

Router 2

10.1.1.0 directly connectd
10.1.2.0 directly connected
10.1.3.0 via 10.1.2.1 hop 1

Router 3

10.1.2.0 directly connected
10.1.3.0 directly connected
10.1.1.0 via 10.1.2.2 hop 1

Router 4

10.1.3.0 directly connected
10.1.2.0 via 10.1.3.2 hop 1

they will keep filling up
Remenber they keep sending updates every 30 secs so in the next 30 secs he will see the router 2 has new updates in his routing table and will sent it to router 1(note that it will keep sending updates even if it has NOTHING new on it every 30 secs)
This is why this protocol blows it uses lot of CPU so a cpu overloading might occur


Router 1
10.1.1.0 directly connected
10.1.2.0 via 10.1.1.1 hop 1
10.1.3.0 via 10.1.1.1 hop 2

Router 2

10.1.1.0 directly connected
10.1.2.0 directly connected
10.1.3.0 via 10.1.2.1 hop 1


Router 3

10.1.2.0 directly connected
10.1.3.0 directly connected
10.1.1.0 via 10.1.2.2 hop 1

Router 4
10.1.3.0 directly connected
10.1.2.0 via 10.1.3.2 hop 1
10.1.1.0 via 10.1.3.2 hop 2


I bolded the ones that in the next 10 seconds had a new update.... the others receive the update but since they got nothing new to add they stay with what they got..

NightShade1

Feasible distance and Advertise distance

this is simple

FD=fiasable distance

AD=advertise distance

AD = distance from the neighbor router to the destiny

FD = AD+the cost between your router and his neighbor

is simple as that
Simple example

R1

---

R2

---

R3

---

R4

Cost from R1 to R2 is 4

Cost from R2 to R4 is umm 8 so AD is 8 cause is the cost from the neighbor of R1 to destiny

FD = 4+8=12

in this example i just gave you a uniroute from R1 to R4 but normally there will be many routes with many differets FD, the router pick the one with lower FD and put it in the routing table.

NightShade1

Variance command!!

Some stuff you gotta understand before this
1-Feasible distance is the best metric along path destination
2-In the topology table you will see all the feasible successors and then it will pick the best one(the one with the lowest metric) of them as a successor and will place it on the routing table, and the other feasible successor will be still on the topology table as a back up route
3-By default, if there are multiple equal-cost paths to a destination the router will load share across up to four paths

Now what we try to accomplish doing this command? well the idea of doing this command is putting others Back up routes in the routing table so it can distribute traffic through these routes.
Now knowing that lets go to the variance thing



E-B-A with a metric of 30


E-C-A with a metric of 20


E-D-A with a metric of 45

the only route that will be in the routing table will be ECA because is the one that satisfy the minimum metric which is 20 and the feasible condition. The EBA will remain in the topology table as Feasible successor.

Now with the variance you can vary the metric condition....
if you set the variance to 2 like in the example then the minimum metric will be 40 instead of 20. Which means that it will accept any route with metric of 40 or less in the routing table... so it will accept EBA also in the routing table to do unequal load and balancing

Remember that it need the minimum metric AND satisfy the feasible condition
if you set the variance to 3, it will accept the metric but the feasible condition will not met so it will be not added i mean for E-D-A

And which is the feasible condition?
the feasible condition tells you that the Advertise distance cant be bigger than the Feasible distance... in this case the feasible distance is 20 and the advertise distance for route E D A is 25....
EDA will be not be added to topology table as a feasible successor because it doesn't met the feasible condition(i think you could still see this route doing show ip eigrp topology all-links

NightShade1

How does the routing table in the EIGRP fills up?



Look you see 3 routes to net X

E-B-A with a metric of 30


E-C-A with a metric of 20


E-D-A with a metric of 45

in the topology table you will see

FD=20
Why? well because the it picks the one with the lowest metric of it to be the Feasible distance

now in the topology table you will see also all the feasible successors
you will see
P 172.16.81.0 255.255.255.0, 1 successors, FD is 20

via 172.16.81.28 (30/10), Serial1 lets say this is E-B-A

via 172.16.80.28 (20/10), Serial2 lets say this is E-C-A

via 172.16.80.31 (45/25), Serial0 let say this is E-D-A ---> this one you will NOT see it on the topology table because the Advertise distance is higher than the FD... the advertise distance is 25 remenber Remenber for being here it must meet a feasible condition which is "the Advertise distance cant be higher than the FD"


now in the routing table you will just see
172.16.81.0 255.255.255.0 via 172.16.80.28
Why ?
well because is the one with the best metric
All the others will remain in the topology table as a back up routes

NightShade1

RIP route auto summarizing(i took an example of a fellow in other forum i was too lazy to make my own heh)

Rip just auto summarize when it crosses a different network boundary which means that if the networks has the same major network then the auto summarize wont happen

The other thing is that they all must share the same subnet mask

So Ex.1

10.1.10.0/24|

---

R1<---192.168.1.0/30---->R2<----192.168.1.4/30---->R3----|10.2.20.0/24

here R1 will sned an update to R2 with Network 10.0.0.0
(As its a cross a diff. class boundary/ diff.network)

Ex.2

10.1.10.0/24|

---

R1<---10.16.3.0/30---->R2<----10.16.3.4/30---->R3----|10.2.20.0/24

here R1 wont send an update to R2 !!!, Its the same class boundary , using the same Major network 10.0.0.0 but the update will be supressed, AS the out going int. have a DIFFERENT SUBNET MASK /30


Ex.3

10.1.10.0/24|

---

R1<---10.16.3.0/24---->R2<----10.16.4.0/24---->R3----|10.2.20.0/24

NOW , the network will work fine, you will see that R1 will recive an update with 10.2.20.0 ,10.16.3.0 &
10.16.4.0 and the interface on R1 thats will recive the updates will apply a mask of /24 on them.

NightShade1

i ll try to make more tech notes another day ill try to make it in a better order and more organized like the ones on the tech notes of CCNA sticky.... But for now i just copy and pasted Notes and stuff that can be useful and ppl have asked me several times, and my answer to them