MPLS & BGP Simplified!

okay guys I am really excited about this video because it's taking me absolutely years to get to the point where I can simply explain mpls in five minutes now a bit of background I have designed large MPLS networks I have managed them and I've always struggled to you know explain MPLS in simple terms how my disapproval itself how is working for a service provider networks in core enterprise networks you know and and this is probably the best explanation you will find on the internet I have I'm yet to come to some sort of explanation which is better than this and what side happening is I start explaining best to people and every time I would explain it in this sort of analogy after couple of minutes people would be like oh it just all makes sense to us you know it would be like a lightbulb moment so that's actually the reason behind this video so without further to do let's 12 into what I'm trying to talk about so let's focus on this blackboard here what is MPLS it's a tunneling mechanism right but what you will notice is a large enterprise or source wider networks MPLS is run along with nigp eius eius or OSPF and BGP you don't just run MPLS on your own and you actually can't run MPLS on your own like this yes you could figure in place and statics but this is a basic five minute video to just get you to understand what is this what is a tracker between the three of them so you have a network on the Left I've actually created a network a diagram to explain so we've got a network of three rooters and we've been told right go ahead and just deliver this core network this will be the core network that is going to interconnect three data centers so in our case we have p1 which is representing one data center p2 which is representing another data center and p3 which is representing another data center now obviously the whole idea of a core network is the core network connects all of networks together right perfect so let's get into the nitty-gritty of course we're going to we're going to run these three protocols web I mean not protocol IGP a kind of protocol a kind of IDP isis or SPF is what we're going to run a kind of MPLS protocol RSVP LDP segment routine what are we going to run so I'm not going to get into exactly the nitty-gritty of that protocol but the terminology so IDP represents eius eius or SPF MPLS represents rsvp-te and LDP second routine and BGP represents you know BGP in our case I BGP in this example okay so what are we first going to do is simply configure the base infrastructure right which in this diagram you can see as a point-to-point I know I've given instruction for just for ease of addressing so this is simply a point of a network point of my network pipeline network and the p1 p2 p3 now have three loopback addresses 10.1.1.1 tend not to door-to-door to 10 or 3 Row 3 Row 3 now the first goal in deploying a core network is to make the loopback speak to each other that is the primary reason why we need an IG P and in our case we don't know SPF or you don't know or we run eius eius we we will end up making the loopback speak to each other in other words P t1 would be able to see p2 and p3 is loopback they will be able to ping them right job done that is pretty much all you need from an IG P in our scenario and and we referred you to do without this nothing will work we won't have any less working we won't have BGP work so the first step is we get our I GP working perfect now once our IDP is working thus done the next step that we need to do is now we can run MPLS now we've simply in our example I'm just going to say we're on l DP l DP runs and all of a sudden we have tunnels enable now what are these tones therefore and from from the perspective of PE one you now have a tunnel to PE 3 and you eternal 2pe - this is an MPLS tunnel and what this means is that to get to the loopback address of 10.3 door 3 2 3 we are going to know what label to push and encapsulate the IP packet inside you know the MPLS label that comes in front of the IP packet so we will now know what label to put in front perfect job done so MPLS starts running and we now have a beautiful MPLS Network running we have we have tunnels everywhere everyone can speak to each other perfect job done but what next nothing is going to use those phones because why do we create these terms we created them for traffic right we want traffic to use these tunnels these are just endpoints and they're like I explained before their whole data centers on the back of these rooters right so when we kind of want those data centers to be able to speak to each other and send traffic right that is where BGP comes in what we're now going to do is we are going to we are going to run BGP between p1 p2 and p3 and remember IDP enabled us to run BGP exactly how enabled us to run MPLS so what we're going to do is we're going to run BGP and we are going to simply and get prefixes across from different isn't it so obviously in this example of this for simplicity remember their incentives but just I've just removed data center that's in between and I'm just product customer here customer your VM just name it whatever you feel like for your ease both this can idea that this is something that is trying to speak to this here a prefix from Google for example so imagine that you ended up running BGP and Big B's running fine you've got it all good image yes fine just leaved I wanted to just share the window screens nicely boy I've just just won't let me do it nicely I probably can try here you know what it's not my thing I'll just ignore that anyway so without further ado Hey where were we right so we ran BGP and now all our prefixes from here and from Google are shared across now what you will see is in all these routing tables on these routers you will see prefixes from Google shared across and for simplicity what I'm going to do is I'm going to give that prefix a name just mess up again there you go right that's just my OCD right ok so where was I I am going to add an option it's going to add a bit of text a prefix from Google so as you know 8.8.8.8 belongs to Google I'm going to add that prefix then I'm going to put it on here in front of Google right so that's a prefix an example that we want and that meaningful traffic is going to go to and this customer is trying to speak to that now as soon as when bgp what will happen is that a dot a dot of 0 is going to get shared across to p1 and p2 perfect job done now how do we start using MPLS remember we built tunnels across right and our tunnels are everywhere they basically here-here-here everywhere and so from a customer journey point of view now that pe3 knows how to get to a data 8.0 any things that to get to a toyota 8.0 in this next hop is 10.2 2.2 what will happen is a packet from here from customers going to come to P 3 and P 3 is going to look at in basic terms obviously in the next video I can go in a lot more detail about I know 3 and all those tables it will probably use before now PE 3 is going to look into this routing table and it's going to see an entry and entry is going to say to get to 8 8 8 0 just use label just push a label for example 5000 1 0 3 and send a across at this interface G 0 3 forget penetrable penultimate hop pop in normally we use the label 3 but just ignore all of that and which which basically is just just ignore that for now because it's not part of this it's not scope of this basic video but while I'm trying to explain here is that it's going to push that push that in that an IP packet push that label on top of that packet I was going to send a hair and when p2 receives that is going to remove that label and it will then just use normal routing in and send traffic forward to Google and return traffic the same way when the return traffic comes back is going to come here and then p2 is going to know the prefix of the customer through p3 although he will have two paths to go through either use that label or most likely will use the label LSB label switch path and through here and it will go back now coming back to what we were talking about before now as you can see that our IGP enabled our loop backs to speak to each other perfect now the loop access speak to each other then we're an MPLS to enable tunnels and then for meaningful traffic we had to run BGP and that is the troika I was talking about ok so back to basics your IGP is definitely needed with our IGP you're not going to get anywhere lgp is what enables loop access speak to each other without IGP your MPLS Network your RSVP LD be sort of signaling protocols segmenting nothing would work then BGP you need ijp for BGP to work because you're trying to bail bgp between loop backs that's not going to work either because p3 has no clue how to get to 10.2.32 remember that's a loopback address airspeed v3 knows about this network only so MIDI P needs ICP as well p3 needs to build a beach been able with this loopback address and this loopback address you can't do that with the IGP so your AGP is your base next you're an MPLS you run BGP what happens both of them start using each other you don't run MPLS you just have an on and core network where you'll just be IP routing you won't be able to use the power that comes with MPLS for tunneling where you can hide so many protocols behind the MPLS label sack similarly without BGP what's the point of this tunnel you can't send meaningful traffic there so it's a tracker that just everything complements each other and that is how simple a core MPLS Network really is so when you can go to any big sort of network and pretty much this is how they're working and I call this video was to was to explain to you how this trike has working if you keep it keep things in mind this way the IDP is your base MPLS runs through ibgp IG p.m. and sorry bgp runs using IG p and they both complement each other that's it job done and that's how simple it really is so when I hope you found this video informative and I'd like to thank you for viewing

Channel: Route My Packet

Views: 3,889

Rating: 4.0117645 out of 5

Keywords: mpls explained, mpls configuration, mpls network explained, mpls vs sdwan, mpls vpn, mpls traffic engineering, mpls network, mpls l2vpn, bgp protocol, bgp routing protocol tutorial, bgp configuration, bgp attributes, bgp, igp, bgp conmmunity explained

Id: FRenSu5GWzw

Length: 11min 55sec (715 seconds)

Published: Sat Dec 07 2019