MicroNugget: BGP Configuration Explained | CBT Nuggets

Start learning cybersecurity with CBT Nuggets. courses.cbt.gg/security
In this video, CBT Nuggets trainer Keith Barker provides a brief overview of how the Internet works together with BGP (Border Gateway Protocol), and how BGP chooses the best route when presented with multiple paths. Learn more about how BGP works to get started on BGP configuration.
We can all give thanks to BGP, which makes important forwarding decisions to choose routes, for the entire Internet.
This happens because entire clouds of networks receive autonomous system numbers (ASNs), and get addresses distributed amongst them.
The Internet Assigned Numbers Authority is responsible for managing all the IPv4 and IPv6 addresses. Regional Internet Registries handle the many requests that are generated by service providers to gain additional addresses.
With those blocks of addresses, enterprises can subnet and rent them out to customers. After a company takes those addresses and makes those routes and subnets, BGP comes in to advertise the routes to neighbors.
Those advertisements get propagated across the entire Internet. But when an AS learns that there are two different routes to get to a given network, it has to determine its path.
Fortunately, the address itself includes how many hops were necessary to reach the end. So the router just has to choose the one with the fewest hops.
0:20: An overview of what BGP does and how it chooses its routes
1:30: The relationship between IANA and RIR
2:35: BGP's job advertising routes
3:40: How BGP chooses a route between options
4:55: Live demonstration of the determination process
🌐 Download the Free Ultimate Networking Cert Guide: blog.cbt.gg/gus4
⬇️ 13-Week Study Plan: CCNA (200-301): blog.cbt.gg/9m0k
Start learning with CBT Nuggets:
• Border Gateway Protocol (BGP) | courses.cbt.gg/gda
• Juniper BGP | courses.cbt.gg/xgq
• TCP/IP IPv4 | courses.cbt.gg/m8f

[@EViL3666]

8 years old... and still the best video I managed to find on this! Great job, thanks!

[@KeithBarker]

You are very welcome. Thanks for your kind words. Best wishes, Keith Barker

[@vuttham6]

11 yrs old video, but still it's the best to video to all the network engineers who just lost their basics and also the best video to all the novice network engineers. Great Job Keith.

[@vuttham6]

Can anyone guide me on the below: Is it worth taking CCIE course at the moment where evryone are moving towards Cloud Tech

[@muhamj]

Keith Barker always makes things easy to understand. I really enjoy his style.

[@dhhsncnd6107]

Keith's vids are always inspiring. I wish he has more detials for BGP in his personal channel.

[@ardentdfender4116]

This is so well explained and explained by Keith. Well explained for me and impending CCNA exam. Very relevant in age of FB down earlier this week.

[@abrahamoz1697]

Thank you very much Keith, great BGP intro video it makes a lot more clearer now... :))

[@xfox360]

Thanks for the easy brief introduction to BGP, still one of the best vids out there.

[@francescogalli4033]

Simply love this channel✌🏻 As junior networker, it's my Bible 🏅

[@MyJonathanAlex]

Keith you are the man...thanks for putting this out there...simple and clear..you rock.

[@fieldzee09]

This is the best video I've seen to explain how this works. I am a visual learner, so this works for me.

[@KennyBentley7]

Excellent video! I'm studying for some IT certifications and have learned a lot about TCP/IP on a LAN, but I've been curious about how it works on the Internet. Thanks for sharing.

[@KeithBarker]

Hello- Great question. If all the items up to and including the AS hops are equal, the following would be used, in this order: lowest ORIGIN type (IGP, EGP then Incomplete, in that order) if that is a tie then lowest MED if that is a tie then eBGP over iBGP learned path if that is a tie then lowest IGP metric to the next hop and there are a few more if those are tied as well. Before the AS hops, there are a few checks it also does. Hope that helps, Keith

[@enriquegabriel7708]

These guys are the best teachers.

[@SkynetWanoG6th]

I Love Keiths teaching style he is the best

[@ryanmyers9092]

Absolutely great video, explained very well easily understood, Way better then the classroom....

[@jsuswaram]

Awesome presentation..Many thanks!

[@rOnNiecHaRles16]

PLEASE TEACH US MORE LIKE THESE, ONCE I GET HIGH SALARY I WILL ENROLL TO YOU CBT NUGGETS I PROMISE THAT!

[@oneznzeroz]

Are you going to yell at your new job? ;)

[@newentshini1229]

@@oneznzeroz YES

[@karimaurelie1]

allways simple to understand with you thanks a lot

[@MuhammedParlakci]

Thanks it was very helpful!

[@MrBruksroy]

Very good explanation! Thanks a lot!

[@TotemYuri]

thank you for the video, very helpful!

[@KeithBarker]

Thanks for your post Pete! Keith

[@phurbalama986]

How about other bgp path attributes?

[@dossosindou7501]

Hello Keith which layer of osi BGP use I am confusing please explain to me

[@KeithBarker]

@@dossosindou7501 BGP is a routing protocol, that uses IP at layer 3, TCP at layer 4, and the updates and message it carries are at the application layer.

[@KeithBarker]

The router can recursively figure out the correct next immediate hop, to reach the BGP peer who sent the packet, based on its routing table and will forward it based on that. If MPLS is used, label switching could come into play, but that is another story altogether.

[@pjtemplin]

MED is (normally) only a factor when there are multiple links from one AS to another AS, and your question only applies if the routers within one AS are running vastly different software versions so as to pick different favorites. The "incomplete" is merely an origin code, not an incomplete ASN.

[@hazgyb]

Thanks!

[@ShriramShastik_Crafts]

very nice explanation...Many thanks

[@dalepeterson9907]

Solid explanation. I dig it...

[@SherzTutorials]

Thank you so much!

[@jimmycheong7970]

Amazing!

[@djbneozen]

Thanks Keith. I do understand that. I was trying to ask what the dif was between the "old" standard and new. "Before the most recent edition of the BGP standard, if an UPDATE had no MULTI_EXIT_DISC value, several implementations created a MED with the least possible value." I was wondering if you could briefly expolain this as I bet there is tons of legacy BGP out there. Also, if BGP requires an assigned AS#, how can there ever be an "incomplete"? Do a video on tie breakers!! lol

[@kosta250]

good explanation!

[@RGOWAL]

Hey guys from CBT Nuggets, How did you program that explanation example you put at the end? I'm currently studying BGP and I want to include an explanation video just like that one you guys did in my presentation. Thank you

[@b1swoll273]

Lots of light bulbs went off with this video. Much appreciated

[@vinayrp3081]

nice explaination !! but does it not consider the amount of congestion in the path ??

[@priyankaaher]

awesome... thanks man

[@XSFlanger]

Good explanation as usual. Can you tell me please, since AS34 and AS56 are transits, they also have to have some sort of IGP in conjunction with EBGP, to carry external routes in the internal network between border routers right? Is it possible to carry those routes with other IGP like OSPF instead of IBGP. For example, redistribute external routes from EBGP right into OSPF and get them back on other border router to advertise to next neighbor AS, or only IBGP will do that?

[@KeithBarker]

The old way, was to presume a non-configured MED was 0 (which would be less than a configured MED of 2). "Incomplete" doesn't refer to a missing AS #. "incomplete" means that BGP is not sure how the route came into BGP (network statement wasn't used). Could have be redistributed.

[@jakedavitt8446]

What of the route was the same size on both of the paths? How does it choose which one is the better one?

[@peterkim1999]

So the *> best path gets populated into the routing table. Now when packet comes in and matches the "best path" it egresses to the "next hop" which is a neighbor? What I am confused about is that what if the next hop is not a directly connected neighbor? In that case, what routing method is used in the transit?

[@kennrich213]

Nice MicroNugget. Anyone know what terminal emulator is being used in the latter part of the video?

[@mariosolano629]

BEST video ever

[@cbtnuggets]

Thank you Mario!

[@thilolg]

Brilliant

[@lottebirrenkoven8162]

thank you.

[@KeithBarker]

If one path has a MED of 2, and the other path has no MED configured, the current standard is to treat that missing MED as the highest (worst) possible value. If that was the case the MED of 2 would be considered the best.

[@pjtemplin]

Lowest MED is (normally) only a tiebreaker if you have two links to the neighboring AS.

[@aniruthoblah]

Hi Keith, thanks you doing this video. I have a silly question. Does the BGP not take into account the time taken between two ASNs? What if the time for the longer path is actually faster because it has Google fiber or some other fast connectivity between them?

[@KeithBarker]

+Aniruth Oblah Time isn't a metric built into BGP (other than a BGP session between 2 routers could time out). Service providers can use other tools, including probes and monitors, to verify and check connectivity, including delays, and have BGP respond based on those other tools. Here is a link to an example of that: linkstate.wordpress.com/2011/07/15/ip-sla-and-object-tracking/ Cheers, and happy studies.

[@lrellisderth118]

Thank you for the video. Is it only the BGP routers in each AS that participate in BGP then? And how many BGP routers do you need for an autonomous system (AS) based on the number of routers in the network?

[@crza96]

Correct. BGP Routers connect AS's together, so for a company it would be the edge router(s) depending on the size. So a company can use Interior Gateway Protocols, like OSPF or EIGRP to route it's own network, and then use BGP as their External Gateway Protocol to get out to the internet. Once leaving the AS, you're traveling on the internet through Internet Service Provider land, which is much bigger in scope.

[@Anabieh]

Thanks

[@sekarjvm]

Hi i have a problem in my epgp with multi isp in mikrotik router. In a day 5 to 6 packets getting dropped from an outside network which means blackout our network users between the duration. What was the issue can you help I can't figure it out.

[@guillermotijerino]

nice excellent video

[@An_Escaped_Mind]

What is the program that you are using to type your commands?

[@moscovciucmircea9039]

GNS 3

[@djbneozen]

"The current standard however specifies that missing MEDs are to be treated as the highest possible value. Since the current rule may cause different behavior than the vendor interpretations, BGP implementations that used the nonstandard default value have a configuration feature that allows the old or standard rule to be selected." - How about this?

[@briankelley8311]

Nice video Keith! Inside each AS, there could be half a dozen hops and increase latency. Does BGP have a mechanism to calculate that?

[@enriquegabriel7708]

No. It is not a link state protocol.

[@nialldonaghy5940]

If you have two peerings with another AS, and learn two routes to the same prefix from that AS, they might advertise different BGP metric (MED) values for each route. Lowest metric wins (normally). They may configure peering A's advertisement as metric = 0 and peering B's advertisement as metric = 100. Your AS will prefer lowest metric, so path A is chosen. Another way to set the metrics is automatically, by copying their IGP metric to the BGP metric. In this way, their internal path differences are reflected. This is a simple but effective mechanism. BGP is not a link state protocol. Metric is very crude and simple; latency and bandwidth are not conveyed. However, BGP communities and other NLRIs and features can and do convey additional information. You might lookup BGP DMZ link bandwidth communities for example.

[@danielsaan1976]

Folks, BGPlay really helps here.

[@djbneozen]

So when the number of AS hops are equal...what is the tie breaker? (surprised I don't see anyone asking this, nor is it mentioned in the vid)

[@scottza]

In case after 6 years this question has been plaguing you here is an answer. How these 'tie-breakers' are dealt with goes down to how the AS has been set up and configured. In general, BGP routing is decided based upon a number of criteria (In order). 1. Local preference value attribute: A policy-based decision is made regardless of the shortest path. If an AS is configured to not favour an AS it will attempt to find other paths at all costs. 2. The Shortest AS-PATH is chosen: If no relevant policies to make a decision, then as demonstrated that the path with the least AS steps to the destination will be chosen. 3. Closest NEXT-HOP router: If there is no shortest AS-PATH, then instead the closest NEXT-HOP router is chosen. The closest NEXT-HOP router will be the closest router of the 'tie-breaker' within the same AS. (This is known as hot potato routing :) ) 4. Else, any other criteria that are specific to the router or AS's BGP implementation. Hopefully, that might help you with your exam ;-) - Although this is likely more relevant to your kids by this point.

[@estee_tey]

​@@scottza I am currently studying BGP stuff in school, your comment helped me! Thank you :D!

[@scottza]

@@estee_tey Glad to have been of assistance. ;)

[@ardentdfender4116]

@@scottza One to two years later, this answer is extreme helpful to me in understanding and for my impending CCNA exam. Thanks a billion bytes!

[@KennyBentley7]

Might make the head of one who's not familiar with the mechanics of IP, but for someone who knows how IP works on a LAN, very helpful. What tool were you using? Did you have a bunch of routers that you telnetted into, or is that like a router emulator or something?

[@xfox360]

maybe its gns3

[@garymcleanuk]

Makes it sound so simple but this is only the beggining. What happens when the path has equal number? When does Local Pref and Remote Pref come into effect? Then theres Tags

[@SM7SM7SM7]

Isn’t the shortest path an ospf thing for the first decision? Bgp first looks for the path with the highest weight, and the shortest path is the 4th decision. Just a clarification

[@sunisilver]

As in show output local preference was 100, and your saying it's due to as path....

[@andn9777]

100 value by default

[@john63us]

Whay about weight, Local Pref, then shortest AS_PATH

[@rahulr9539]

Which software was that ?

[@steezey]

This is not entirely true. It does not necessarily look at just the shortest path. The BGP routers look at whether the path came from a customer, a peer, or a provider. BGP routers will avoid using providers as it costs them money to use, and will not advertise routes that came from a provider to other providers or peers. For example, a peer can take advantage of such a path and use the route though the advertiser because it will not get charged for it, causes the advertiser to get feed.

[@andn9777]

Yes you are right but this is how bgp algorithm is working by default...

[@davidratkay9281]

Also It will look at the LOCAL_PREF before the shortest path, which means it chose the route with the shortest AS path BECAUSE OF LOCAL_PREF in this case.

[@Alireza-zc3kc]

Thanks #kieth_barker

[@davidratkay9281]

It looks like it chose the shortest as path route BECAUSE of LOCAL_PREF not because of the shortest AS path.

[@mattjoe67]

Hello Keith, thank you very much for clear explanation, Could you please consult Mr.Jeremy how to teach or train him how to explain It, I listened to all his nuggets audios for almost 7 months, I couldn't understand him and DOES NOT KNOW the method to explain these It subjects clearly, he always around the bush/confused not coming up with a clear explanation, he is just goofing, I am an instructor for almost 15 years, please let him know, he needs to get his act together.

[@AdamDavid]

Hahaha a /12, Sure!!! LOL

[@mattjoe67]

Very confused, Just explain the BGP functions, don't be around the bush.