BGP Peer Groups

I spent the better part of my morning learning how to configure BGP Peer Groups. Peer Groups provide a way to save router processing time by not having to send an individual update for each IBGP neighbor. If your Autonomous System has a rather large amount of routers (neighbors) running BGP, the processing time each router would have to run could be pretty intensive. Remember that when running IBGP within a transitive AS, all neighbors should be fully-meshed with one another due to the way BGP sends updates (unicast). I was also introduced to distribution lists and how they could be used to tune BGP updates. In the lab I created I used a distribution list to permit internal IP networks from being addressed outside the AS to other AS's.

As you see in the Router A BGP example, the commands to create a peer-group, distribution list, and creating updates for specific networks can be quite extensive compared to other routing protocols.

BGP Neighbor Configuration

After reviewing my CCNP study guide book this morning I finally ran into the material that would of helped me out tremendously yesterday. Sometimes I get a head of my self when studying because I will lab or study topics which hasn't been presented to me yet. This can be good or bad depending on how you look at the situation. What I came to find out when configuring BGP is that there's a huge difference between IGP and BGP network commands. When using the IGP network commands you are determining which interfaces should be apart of the IGP routing process. When using BGP network commands, you are determining which routes in the IP Routing Table should be apart of the BGP routing process. It's also important to note that the network you are referencing in the IP Routing Table will have to exactly match in your BGP network command.

For example if you have the route 192.168.1.0 /30 in your IP routing but type the following BGP network command:

network 192.168.1.0
or
network 192.168.1.0 mask 255.255.255.0

This will not work you will have to type the following:

network 192.168.1.0 mask 255.255.255.252

In order for BGP to include this route in the routing process.

Configure the Network Free Lab Book

To all of those who haven't signed up to the mailing list, I wanted to let you all know that the free lab eBook should be complete within the next two weeks! The people who are signed up to the mailing list will be the first to gain access to the download once it's complete. You can sign up on the right side of this blog, I try to send out a news letter weekly.


There will be two versions of the CtN Lab book. The free version will be based off of mostly ICND1 material. The full version will contain the following:

-Working Solutions for all labs

-More than twice the amount of labs offered in the free lab book

-Both ICND1 and ICND2 material (Entire CCNA Exam Blueprint)

And many more features to come! Pricing has not yet been determined for the full lab but I want to make sure that's it's more than affordable to everyone. If you have any questions please feel free to email me at: configthenet@hotmail.com

Configuring BGP

I spent most of my early afternoon configuring BGP for the first time. I was honestly a little nervous about if I was going to get this protocol up and running because I hear so many stories about how in-depth this particular protocol was. It actually took me a little while to get my lab up and working with this protocol because I wasn't sure if i needed to synchronize my BGP routes with my IGP (EIGRP) or not. I ended up using loopback interfaces as my update sources and configured my AS 65101 as a full-mesh BGP even though I' sure I didn't need to in retrospect since the AS wasn't a transitive AS.

Transitive AS's pass on BGP routes from one AS to another AS, in other words it's like a middle-man in the BGP routing process. I also had problems with my network statements, at first I didn't realize you even needed to use network statements since you were using the neighbor commands. Even after using network commands my BGP routes still wouldn't propagate to each other. After a little research i decided to define the specific subnets that I wanted to be used for routing updates. As soon as I did that all the routes came right up!

BGP Attributes

Today I studied the different BGP Attributes that are used to determine routes. There are Well-Known Attributes and then there are Optional Attributes. Well-Known attributes must be recognized and propagated to BGP neighbors. Optional Attributes may be propagated to neighbors depending on the attributes meaning. Within these two types of attributes, there are sub attributes.

Well-Known Mandatory Attributes (Must be included in BGP Updates):
- AS-path
- Next hop
- Origin

Well-Known Discretionary Attributes (Not mandatory to be included in BGP Updates):
- Local preference
- Atomic aggregate

Optional Transitive Attributes (Must be passed to other AS's even if attribute isn't used):
- Aggregator
- Community

Optional Nontransitive Attibutes: (Doesn't have to be passed to other AS's):
- Multiexit-discriminator (MED)

I've also reviewed how BGP Synchronization works along with the many BGP message types. Synchronization tells BGP to wait until all routers have the same IGP information before updating other AS's with the info when using redistribution of BGP into IGP. From what I've been reading synchronization is outdated as most BGP updates are to large to be redistributed into IGP's anyways (scalability). BGP message types are used for establishing BGP neighbors along with providing keep-alives and BGP router updates

BGP Path Vector Characteristics

I spent my morning learning about BGP and its path vector technology. Path Vector is like an updated scalable version of distance-vector routing. However it has one main difference than any other distance-vector or IGP. BGP doesn't use broadcasts or multicasting since it uses TCP as it's transport protocol. TCP is used because it's able to send a large amount of data reliably, the BGP table has over 190,000 entries in the full routing table currently! This size would actually be in the millions but thanks to technologies such as CIDR, the table size has been reduced considerably. Since BGP doesn't use any type of broadcasting mechanism, an AS being use to transit BGP routes between other AS's must be fully meshed within the transit AS. In other words every router running BGP in a transit AS must be seen by every other BGP router in that AS. this is becuase the BGP updates are sent as unicast messages which aren't forwarded in a multicast manner i.e. one-to-one mesages.

CCNA Configure the Network free lab book is on the way soon! I should be completed with all of the labs for this eBook by sometime next week. I will also need my work to be edited for the final version as well. If you haven't already please subscribe to the mailing list on the right to be the first to receive updates and a link to download the free CtN CCNA lab book!

CCNP BGP Overview

Well today I finally started the chapter for BGP, I've been waiting to get to this one for a while. I went over the first section of this chapter which gave me a brief overview of what BGP is. I can already tell that there is a lot to this protocol looking at the number of RFC's alone! The fact that this is the biggest chapter in the book (over 100 pages) is another reason. I ended my study session by viewing a few multihoming options an AS has when receiving EBGP updates from multiple ISP connections.

Don't forget to sign up on my mailing list to the right to be the first to receive my free CCNA lab book! The book is only a few weeks away to being completed!