360 pre-assessment learning :(

Well, there were quite a few areas left for improvement after the pre-assessment.
I failed misrably in route summarization for all routing protocols :)
So, I figured since that stuck out so badly, I should focus on it.

So today I turned on GNS3, and played with it. On Router 5 (R5) I created 4 networks using loopback addresses.

20.20.1.1 /24
20.20.2.1/24
20.20.4.1/24
20.20.8.1/24

Advertised these networks into OSPF
router ospf 1
network 20.20.1.0 0.0.0.255 area 1
network 20.20.2.0 0.0.0.255 area 1
network 20.20.4.0 0.0.0.255 area 1
network 20.20.8.0 0.0.0.255 area 1

Now the fun part. I thought ok, time to summarize, do the math:

as the 3rd octet is the focus, I just did the math there.

20.20.1.1 00000001
20.20.1.2 00000010
20.20.1.4 00000100
20.20.1.8 00001000

as we are already using 16 bits, (20.20.x.x) we count the remaining like bits to determine the mask to use for the summary. So we see we can use 4 bits, as the 5th bit in this octet changed. This leaves us with a 255.255.240.0 or / 20 Mask (16 bits used for the 20.20 part of the network + 4 bits in our 3rd octet)

So the summary would be 20.20.0.0 /20.

Next I applied the command using the summary-address command under the RIP process. THis did not work, But I was like how , I just did the same test in RIP.

A good friend reminded me summary-address is used for sending to an external AS.
SO i changed the to use the area range command.

Router ospf 1
area 1 range 20.20.0.0 255.255.240.0

Ah.. yes it worked. I apologize for not providing the working output, I closed GNS3 to quickly.

What I did find was interesting. When testing show ip route for my subnets, I found this summary also encompassed networks , that I am not currently using. I did show ip route 20.20.9.0 all the way up to 20.20.15.0 and my summary worked for these addresses even though I was not using them. Interesting huh?

Ok So i was thinking how is this happening? In reading more on this, you can summarize any address as long as ONE of the addresses is in the current routing table. Ah! ok. but why did the summary work for all of my addresses up to non-existing address fo 20.20.15.0, but not work for 20.20.16.x ?

Still Was not sure until I looked at the binary.


20.20.1.1 00000001
20.20.1.2 00000010
20.20.1.4 00000100
20.20.1.8 00001000

It only summarized up to 16 because of the subnet Mask boundary defined in by my mask

The 1st 4 bits, help define our mask, leaving our remaing 4 to be covered in the summary. If you add 1111(15) it equals to 15, hence not being able to encompass subnet 20.20.16.0 in the summary as its across the bit boundary by one bit as I defined in my mask.