Networking for Cloud Computing | AWS BGP Live Training (Live Cloud Architect Skills Training)

welcome everyone this is michael gibbs how are you all today if you're awake and alert and oriented and ready to get started and learn some networking type networking in the chat box in the chat box so we know you're here everyone this is michael gibbs how are you all today bear with me i got some noise in the background i'm trying to monitor what's going on so i can see when people are calling and asking questions so bear with me for having my speaker anyway type networking or cloud networking if you're ready to be ready to go today we are going to have a fun time everyone and welcome and networking is near and dear to my heart from that time back in 1998 where i bought a house full of routers and servers and started building networking to my 25-year architecture career networking has always been at the center of it if you don't have a network you have nothing and you know here's the funny thing about networking all these decades i've been involved almost no one thinks about the network until it doesn't work and then when it doesn't work we go straight to the network so what is a cloud it's a virtual network and a virtual data center so you want to be great at the cloud you've got to get your grade at the network we also got to get your grade at the data center and to that end we're going to be producing a tremendous amount of networking training for you networking is the cord anything we do now in order to do this we're going to bring you some experts from around the world on the networking front you're going to see a lot more of imram and i i want to let you know the kind of background that imran and i have so you can understand where we're coming from both imran and i are both cisco certified internet experts for a very long time we add up our time and experience it's over 50 years worth now imram and i come from a slightly different part of the same career we both started as network engineers i enjoyed my network engineering career more than you can possibly imagine and then moved into architecture and i loved architecture and as an architect i was doing more of the business workflowing more of the systems design and you know who made sure our architectures worked it was the network consulting engineers at cisco people like imran were the people behind people like me making sure that everything we designed worked they would test it they would proof of concept they would show it to the customers i would go to someone like imran and i would say imran look at my design i worked on it can you find any gotchas and imran would look at it and some would say something like yeah over here this will work but it would be better if you tune it this way that's part of the architecture process so i am beyond excited to have imran with me and around i'll have him tell you a little bit about himself but he is a real network engineering expert and a cloud expert and because of this you're going to hear a lot more from imran and i coming soon we're going to have a series of video networking training today is going to be fundamentals and when we say fundamentals understand that our fundamentals are far more than aws advanced networking today we're going to build a network now previously speaking we designed a network if you go to the previous week amaranth and we did the plan we talked about what the ip addressing would look like we talked about what the routing protocol would look like now today we can actually build it so when we build this we're going to do the configuration of the cisco routers imran's going to introduce you to the simulators that we used you're going to hear how and why things work and then we're going to build them and then when we build them we're going to show you routing tables and then we're going to trace packets through the network and then we're going to show you how we engineer traffic from point a to point b we'll build a high performance high availability network then we're going to connect this network to the internet or at least the pseudo internet we're going to see how routes are going to be coming in and after we do that then guess what we're going to connect it to the internet service providers and see this then we're going to connect it to the cloud with two different corrections direct connections and heck you know what we may make an ip6 tunnel to show you whatever pian is but it's going to be a fun time so amram you want to kick it off for two minutes and tell people about some of the amazing tech things you use on it's cisco and juniper as well as the last 25 years before we begin yeah thank you thank you mike for a nice intro i've been in the industry as mike described for almost 20 years uh being architecting designing implementing a very large mega sites project in cisco juniper hammer shroud switch data center ccie i'm here to produce some content for you guys have some poc and testing uh how the routing the bgp works especially in the cloud area so so i'm here to help and we'll have some demo uh uh what's what to expect on this demo uh we laid out some topology for you guys and we step by step we configure the routing protocol or describe what tool we are going to use and step-by-step approach from ebgp ibgp how the route works how do you simulate the cloud in that in a different aspect what are the what are the key concepts are what you really need to understand and how easy for you to simulate any kind of network either in your test lab on a production lab and the tool we use is very handy and and classic mic amram thank you so much great overview so we're going to be playing with a few technologies today so i'll introduce them before we get to work the first thing we're going to be playing with are routers and routers our network devices that connect things from a layer 3 perspective they move your data through the network the way they move your data through the network is as follows they have something called a routing protocol and we're going to configure some routing protocols today a routing protocol is basically something that builds a map of your network if you need to go from point a to point b on your gps make a left here go straight make a right that's your route guess what routers do the same things so internally inside of our organizations we use something called an interior gateway protocol we're going to configure an inferior interior gateway protocol today an interior gateway protocol is something you used inside of your network to recalculate your routes and it does it fast and efficiently and if you've got a cable cut it routes around it very fastly we're going to configure an igp or an interior gateway protocol and it's going to be ospf or open shortest first path a link state routing protocol which is one of the most common ones used in the enterprise environment there's a cloud architect you need to know this and then we're going to configure an exterior gateway protocol which is going to be bgp and we're going to use this to connect to the internet service providers and of course we're going to use this to collect to the cloud or two clouds which we probably will do then we'll manipulate our traffic we'll do some filtering so it's going to be really fun so what we're going to do and how we're going to use this is we're going to use a simulation software now when we train network architects and network engineers we love using the simulation software it's called even g and of course mrm we'll talk about that in a minute but what it's going to enable us to do is build in real time networks test configurations eb eng is used by a lot of organizations in the proof of concept stages they can make it up they have a plan they can build in this environment and then they can test it and then that's part of a proof of concept and a simulated environment which is extraordinarily good also people use this to pass their ccie exams and their ccde exams and of course you're dealing with two network architects and engineers with 50 years experience and three ccis you might see some network training coming from honest really big heavy duty training like how to become a network architect from scratch so lots of fun coming in there but we're going to use ev eng so imran you want to talk a lot a little bit about evmg where you have it set up before we begin yeah yeah you want to let me let me share some topology mic and then i describe what's the uh are you seeing my screen mike um absolutely okay so basically uh guys and mike this is like one simple topology we made for you the the tool i'm using at mike described a ebng and navy ng just describe few things or you can simulate you can simulate basically anything and you can do a certificate testing and have all kind of product is available these are the the routers we are having cisco routers and you can see if i click on the thing i'm on the on the device right now so this is heavily used in the industry for testing and organization use it uh heavily in different simulation uh they are basically is a topology and you can drag and drop any kind of notes uh here and you can have any kind of networks here in the cloud thank you guys sorry mike that's all right i almost worked for aruba many years ago but then cisco gave me an offer at the same time so i turned down aruba well they were some really great wireless technologies it could not go to cisco yeah and of course you've got this running warrant of servers running on bare metal are you running in a virtual machine uh i'm running on the bare metal server and i'm running this uh this evmg as a vm on side to that and as a as like a vm running on a virtual machine uh but i'm running esxi on top of it and running vm for that so there was a lot of the let me share the screen like the features you can have a topology design you can import export configuration you can have a picture import click and play you can have like custom level 2 protocols cpu watch docs multi-users different kind of features you can have it and basically you really simulate the network fully supported and you can run simultaneous labs basically and it just drive from ubuntu uh 16.0 for server from the long-term support and it supports unite unit labs and you basically the ability to learn and transform the real network uh very quick and you can as mike said you can used to i'm still using it for my cci lab scenarios and customer simulations and any kind of network you can simulate so that's a absolutely awesome tool uh other advantages you can have multiple users you can have control permissions you can have a create task laptop scenarios and there is a there is a two kind of versions available i'm using a uh i'm using a community where i'm using a pro version uh the community version is free of cost you can share the design share the design labs with friends and others and import export config and much much more and of course if you go for pro version which is like hundred dollars you have other features are available extensive features are available and you can have 1024 notes available for lab so extremely useful extremely robust and very easy user friendly so let me clarify imran what you mean by nodes he's saying you can literally have a thousand different routers operating all at the same time so that's how big you can actually test these things i want to point out one thing that imran said because i think it could be really helpful to you when you guys build your test labs obviously if you're our students in our cloud architecture real development program you have access to our servers but if you're not with us you know how we always say get yourself a 16 core xeon server with a minimum of 128 gigs of ram and either an nvme drive or three ssds and raid zero what am said is he's running vmware esxi so he's installed vmware esxi on the server he's got this really great lab simulation running on the server and guess what he could still in another vm set up microsoft active directory and in another vm he can actually build an openstack ansible cloud so he can do all this on a single machine through the use of virtualization so i just wanted to point that out i'll go back to you imran yeah so so if you see the topology i'm running these are the all cisco cisco devices uh i'm simulating here is like two or two routers where they're running between ebgp and this like exterior gateway protocol i will talk about it uh what is it and you can simulate this um as like any any of the cloud here like maybe aws cloud you have evgp connection and other thing you can have ebgp with another tcp cloud and other thing is i'm simulating on the left hand on the middle uh we are all running ibgp and i'll explain what ibgp is between ibgp and ebgp and on the left side uh i'm running a dual link for another service provider maybe your internet or maybe another another company you required and how to when you have multiple links on the internet you have multiple links or any other any other company as a provider yeah so that's the perfect apology today we're going to pretend as follows this little blue thing in the center where it says autonomous system 200 that's our customer data center for what we're working with today now the two lengths that imran drew to autonomous system 100 as100 we're going to view that as the cloud provider and we're going to have a primary direct connection to the cloud provider the ethernet 0 0 and guess what we're going to have our backup direct connection to be ethernet 0 1. so that's going to be your high performance high availability routing to the cloud of course as mram said up top as500 we're going to connect to the internet we're going to use bgp um he'll explain all the reasons why we're using ib gpn the way that it's set up when the time occurs and of course we're going to connect to another cloud call it azure on the right side as600 so multiple direct connections to multiple cloud providers and internet service providers high performance routing this is a real network everyone so i want you to understand what it is we're building and then how it works so internally all organizations will have their own internal network and like i mentioned they're going to be running their own routing protocols so they can find stuff that's how they build them out and they're going to be connecting to the cloud too you know what if you don't have a path to the cloud or a route to the cloud you can't reach it so amram's going to do some tuning on the cloud he's going to set up the connections and at some point he's going to show you how we traffic engineer or prefer one link or another he's at some point going to connect us to the internet out of second clouds it's going to be fun so emerald back to you yeah so basically the purpose of uh border gateway protocol basically is having two different artillery systems in two different companies and within the ais within the as you're running ibgp so i uh within the asu run igb and uh so so ebgp and ibgp uh how we redistribute if you're running ospf here to igp uh we'll we'll do that and how we advertise and uh on ibgp and what is the difference uh what are the advantages uh bgp is extremely ebgp is the main protocol exterior gateway protocol and ibgp basically is helping helping to have uh helping for for ebgp to make it work basic basically because if you are if helping to reach the the edge network the ibgp because if you redistribute from ospf to ebg to bgp uh and and from from here to here and how it switched to the to the you know edge uh it will change the attribute for bgp that's why we run ibgp so i i must say like avgp is the main main protocol main routing protocol and ibgp is the one who's helping it so i i must call it like ebgb is like a bad man i ibtp is just helping him uh to make it work so that's the that's the key thing is might wanna add something yeah thank you for doing that so when we're dealing with cloud the cloud is external to us they are external entities and because they're external entities that's why we're using this exterior gateway protocol or ebgp so when it comes to the cloud it's 90 ebgp what ibgp does which is really helps ebgp is let's say i've got a lot of routes to a destination let's say for example i connect to ten dinner for neuron service providers i've got three quarters a million routes from ten different internet service providers if i want my network to have the intelligence throughout the entire network so that every router in your network knows the most efficient path on the way in and the way out you have to run something called ibhps and that's basically tearing the internet's routes inside of your organization why can't you do that with ospf or eigrp or intermediate systems to intermediate systems here's the reason they're not meant to scale to that many routes so what you have to do is you have to take all the routes out of bgp you'd have to do something called redistribution which would be taking them from bgp and putting them into ospf and then you have to define metrics with them and then your ospf routing table will swell with all these what's called e2 external type 5 lsas and what's worse about it is every time something happens in your network you're going to have a recalculation of every one of these routes which is going to sing your cpus to the moon your routers will crash so ebgp connect to external entities ibgp to help it out along the way igps for network layer reachability to make pgp work and you're going to see that when imran does it if you've got no ip connectivity you have no bgp so i've seen people that were trying to write cloud networking books that said bgp is the only routing protocol in use today there is no bgp without network layer reachability so obviously there's more than one just protocol so amram is going to help us do this we're going to start configuring this when we do this by going to the network so we're going to first start with as200 now imran did you design ip addressing schemes for these things yep yeah it's already in here they're already put there okay that's totally fine yeah so i'll i'll explain what is it just for easy for that we used to do this in our cci lab so just to reiterate my case ibgp because if the if you're learning something from as100 it's coming to as200 and it's we need to reach from this edge to this edge that's what you need ibgp because when the routes coming in from es100 you have a you have a as number and the as path is used for loop providence so if you change the as number it will it will change the attribute if you don't use the ib gp so from here to here because you learn something from here to here and from reaching to the edge you need ibgp and of course ivgp is is overlay protocol you need to have igp learn to exchange the information and and it we are talking about routing protocols opening protocols is about uh what who is doing the exchange not routing routed protocols because people got confused the routed protocols are ip ipv6 apple talk uh ipx and spx and routing protocols are ospf isis rip and bgp so perfect so you know one thing before we move on to the configuration i want you to notice this so this is real internet routing that we've set up here so let's say on the left of your screen you see as100 and on the right of your screen you have as600 for as600 to reach as100 it's going to have to go through as200 does everybody see that so mrm will show you um if he points his mouse to as200 which is kind of that like cream can color on the left that needs to send routes to as200 which is kind of a light blue in the middle and those routes need to go to as600 and if as600 passes its routes to as200 and as100 knows the nodes those as six hundred routes through as200 guess what as100 and 600 can talk to each other through as200 now if we want this to work this is fantastic and this is why we use bgp but what if we just wanted to learn the routes from as2 100 and we didn't want the rest of the world to come through us if we passed as100 routes to as200 and 200 exported into as600 as600 to reach as100 but if we didn't want as600 to reach as100 all we would have to do is take in the routes from as100 into 200 and not export them so there's the concept of the concept of no export if you don't export rounds that you learn from someone else you can you won't become transit for someone else so when you're dealing with aws for example are these cloud providers they know this they also know that not everybody's a bgp expert so if you don't know bgp and you take in your routes and you export them to someone else guess how what happens typically speaking the company becomes transit for the entire internet and aws knows this so if you peer with one vpc they don't pass your routing information to the next vpc because they don't want you to become transit for the whole internet so just understand those kind of things while we're doing things so that's why in aws for example they've got all this non-transitive routing and how do we break those rules we use things like cloud hub or transit gateway which are basically route reflectors which we'll show you later but understand the difference between transitive routing and non-transit running as it stands here right now we have transit of routing because this is normal bgp routing so amram if you want to try and go to as200 maybe show the people what these routers look like we'll show how we configure them i may i'll explain what you're actually doing unless you want to explain what you're doing let me put the basic config let me restart the routers from this let me show the people how we can wipe the the device and for the device just so it's so easy if you can see if you right click on it and you can wipe it and you can start it uh once i wipe it out i'll start off all these devices so so much less work in a simulated environment to do these things in a production environment yeah so let me connect to these devices i think i already have r1 while i'm setting up this topology anybody have any questions for me oh looks like you're probably almost ready i'm right now no one second like okay so let's do this once we once we've got the clear routers let's walk everybody through the basics of what they see on the cisco router absolutely that's what i want to see bear with us we're building stuff real time so when you build stuff real time it's always a fun challenge but i'm and i've been doing this a long time so we'll have fun with it these routers do not boot themselves up in three seconds yeah well it floats up very fast so they probably move faster here than they would in real life because got a much bigger cpu if yes we will leave a recording available after the session um definitely it's this boot up like like this mode you want to enter a initial configuration yes and no say no this would be like an auto questionnaire it would walk you through things nobody uses this so vikkas what you have to do when you get these kind of things is you have to have the ios images to put there basically they're like little mini virtual routers and they use the cisco operating system and that's why they're used to do such a good job for testing because it's the actual actual ios operating system and yes it was called ios about two decades before apple decided to use the same term so so don't worry about this rare thing and that's just just using a software to simulate better and you don't have to worry about it's not errors from the router so bear with me michael take your time you're doing great i want people to get to see how all this is built when you get in there are these things you know i want to walk them through what you initially see in the console report before you type enabled um and those kind of things just the people that are not used to working on cisco routers every single day kind of get a feel for it i'm not going to say the cisco router command line interface is necessarily the most intuitive even though for me feels pretty much normal because i've looked at it for so long yep that's all we do every time so yeah pretty much now there are ways to do some of this stuff with api calls but when it comes down to it for some of this stuff we're just going to be getting on there okay so wonderful no are you all wiped off uh i think it's just r1 and r7 the name of the simulator is e v e and j there's a couple of other good simulators that are out there cisco's got a viral one that they use there's another one is it called gn3 or is it gnu3 gns3 gns3 that's another good one i've used with some of my students but this eb eng is the most fully function full capability there's a freeware version which you can use that's quite excellent today we're using the professional version and the reason we're doing it is amram and i do ccie work or ccis and we get pretty deep on these things plus we've got students that need to train on networking because we have students need to train on networking we need the industrial grade environment so you know realistically speaking this is what we're talking about and then when we install things like loopbox we'll talk about them we're going to make everything in these routers clear to you what we're doing so while we're erasing router config does anybody have any quick questions see we're going to brand new routers over here no configuration and we're going to do it from the beginning to the beginning you're more than welcome mr andre but if you've got more stuff anybody's got any questions while you're out there waiting for any room type networking in the chat box so i know that you're not asleep jonojo i'm not sure what you mean by cisco cml2 um but if you want to spell it out happy to help in any way so these all all these routers are are up uh let me show you the original before you type enable yeah can you go back to the base mode okay so when everybody logs into a cisco router here's the first thing that they're going to see they're going to basically if they're plugged into the console port they'll see something called console and it'll say return to get started now i want you to understand when you start out you're going to see router and you're going to see that greater than symbol that greater than symbol basically means you are there with basically user access meaning you can do a little bit of looking and showing a little bit of things but you can't actually do stuff so that's the difference you can see stuff but you're not going to be able to configure this so the next thing imran's going to do when he wants to be able to manage this or manage his router he's going to type enable and enable is going to give him privileged executive mode or basically root access to the router you want to type enable and then what's what we'll watch it's going to turn from that greater than simple to account now what you can actually see is you can see the software version so imran will go back to you and this is going to give you a lot of information you're going to see which version the operating system was being used which is 15.4 oh my god when i started i was starting at versions longer than 10. i think it was 9.3 or something like that so we've come so far over the years you can see when the software was compiled on october 8th 2015. you see the router's been up for three minutes and you can see the warnings about uh cryptographic things and it shows you exactly where the routing image is you know it used to be in flash now you notice you've got unix and linux systems so operating systems a little different so there you go kind of right yeah it's exactly what we're doing here we're paving the way so now i'm showing what's called the running config which is what's actually going on on the router so if you type show run or show running configuration you're going to see some defaults first thing you're going to see is the version and then we're going to see you know different things related to passwords and encryption for example we're not encrypting our password here yet the version before this here you also talk about certain timestamps and certain kind of like syslog messages that you're going to get then um it says no a new model which basically means we haven't started any kind of authentication authorization accounting yet tells us which time zone we're in and regarding snmp that's for management simple network management protocol and then you can see that um we've enabled something called staff which is just a type of switching in cisco um and ronald keep going okay now we've got interface so what is a router it's a computer with a bunch of arms on it send your traffic in this direction this direction out your interfaces which are like your arms so these have got some notice by default out of the box the interfaces are all closed meaning if you plug the cable into it it's not going to work meaning it's secured locked down and it's got no ip address which means it won't work so in order to use these emrem's going to have to put an ip address on them and turn on the ports which he's going to do but i just wanted to walk through the defaults so we've got no web servers turned on by default the line con is related to the console port is if you plug direct in line vty04 is related to telenet for example line aux these routers typically have an auxiliary port and you know you can call a modem you know one of those old things that uses a generic phone line and it negotiates a thing and then you can be plugged in and configuring these things from thousands of miles away i did it all the time why do we do that if you've got ip networking to it you can tell network ss if you've got no network you can't configure it so replace the modem call the modem on the other end of the line if you can even believe this anymore and then we're on the port and we can figure once it's got ip address we switch over to ssh yeah so just to add another point here the the tool you're using for ssh is secure crt uh yeah so it's about secure crt uh that's what i'm told i'm using you know i have a license for it and that's what i'm using what there are other free tools you can use for party and other cli coming with the with your with your laptop so normally typically party and as uh security is widely used i i like the the secret that i've been using for more than 10-15 years old okay okay so just to understand the addressing scheme uh here mike so maybe you can add here so uh i have a r1 like r1 having a 192 1.17.0 network and when it says this is a zero network so the the ip address for uh will be on uh will be on 17.1 here on r1 and 17 dot uh of course this is like a totally different second but that would be the other typical thing 17.1 uh will be the ip address uh for that network and that there are a few loopbacks uh um i'm advertising because of we need to do that and if you go on the the right side this is uh this is the the network you're trying to reach uh we'll add some static routes uh using this to reach that that loopbacks if you say like it goes in the top one like 192 1.34 net 35. so 35.5 will be this interface and 35. uh three will be this interface and network we have a connectivity so so similar fashion on the right hand side 1.46 46.5 will be r5 46.4 will be the uh r4 excellent so one quick thing when emre mentioned the word loopback so with the router you can configure a logical interface it's just logical it doesn't really exist and typically speaking we create what's called a router identifier and it's the loopback address so we basically pick a an address it's typically a slash 32 because we're not going to have any hosts on that submit it's just that thing and we use it to basically promote reachability between ip addresses so that's what we're doing and it's an identifier each router has got a loop back so imran back to you okay so first of all uh we'll we'll configure uh this edges rt and r5 we'll put basically the basic config in all the devices with ip address direct connected ip address and the loop packs for r1 r2 r3 r4 r5 r6 and we will we'll put over there and we'll paste it and once we have the basic connectivity and enterprise are up we will be configuring this evgp here ebtp here and of course dbgb and the multi-hop and then once these are things up and running will configure this portion so we go step by step ask any question if you have anything to ask so i'll i'll go i'll go for r1 uh before you do r1 chin 10 had a quick question if you want to go back to the picture real quick yeah not to not to get in the way of your flow because you're doing so wonderfully and i'm so grateful so what chinton said look i'm not necessarily sure where i should do things i came a few minutes late so i just want everybody to look at this in case you join late i know on youtube people pop in and out and that's great autonomous system 200 in this environment is going to be our data center so our organization's personal data centers can be as200 as100 is going to be the cloud that we're connecting to and that's what we have here done at lynx as500 is going to be another internet service provider and as600 is going to be a second cloud that we're connecting to so now you know what the architecture actually is for those of you that came a little late i don't want anybody lost so we'll talk about why we do ebgp and ibtp when the time comes but everyone's going to go back to setting up the first router so let's build it uh let me use the um the notepad uh notepad is great yeah so because uh you can you can have a copy and paste for that when we build the the documentator and when you type these things i'm going to explain what you're doing post open the host name then the host name is going to be the name of the router that's going to be identified as right now he's basically creating a logical address called loopback so any he right after that he's going to put the ip address for it he's going to have an ip address for the loopback there you go so then loop back zero and then see when you've done it this long we literally type our configs literally in notepad ahead of time right because we don't have to type the full word it's basically going to do this i i have i just want to show them i have to have the country you know i just want to show them the audience okay then i'll turn back to you yeah i want to get in your workflow you're doing really fantastic yeah so basically i have uh these are the interfaces i have a blue back zero i have a loopback one i have loopback 99. and of course uh if you see the i have e00 uh p00 will be the it will be the yes v0192 uh 1.2.12.1 uh 12.1 e0 i think that's just reverse right here e0 how old is that is this kind of what router is this we have we have a gig interface as well so it's supposed to make it make it simple understand the understand the terminology so i was just asking the question we used to have a interface e0 which meant 10 megabit ethernet fe 0 0 it's been 100 gigabit and then ge00 which meant gigabit ethernet so just shows you how long i've been playing with the stuff watch what he's doing and watch what it's going to change now some of these interfaces are going to come right back up some of them might actually have to be turned on oh that's new you used to have to exit out to do showpa ip and brief yeah no it's been a while yeah so you can see the e001 and guru pak i have three loopbacks blue packs it's just a networks and we use it for some other purpose for advertising some to basically now do save so right now so right mam basically means save the configuration could also be the same as copy running config to startup config but we all type right now so right now all we're doing is we're going to put the ip addresses on these devices and after we do the ip addresses we're going to set up some routing but we need basic ip connectivity first absolutely so i do have a like if you see on here hard to have this interface there are so many interfaces so let's put it together i have to have the configuration all the connected interfaces okay now look at this real quick these are the actual interfaces that are up on the router so note every interface has to be on a different subnet which they are also note that it says up so when it says status up it means the interface has been turned on and when it says protocol up it means the data link layer is up so this is completely up you know amram i'm watching you configure these routers and it's making me feel like i'm 20 years old again so we go type something i can think or something makes me feel young i always use notepad so you can replicate and changes i highly recommend you no that is the right way to go yeah because in ccie we have to do so many devices at the same time so and and this is time sensitive as well and you need to replicate a lot of things and even on the production side if you do some some errors and issues you can do that as well so always better to have your configurations planned your ip addressing done when you get there where you give your configs to an implementation engineer whatever the case is plan it out ahead of time now the networking you know look it's not like there's a pretty management console or something like this so you really got to know it in here before you go somewhere else any question and uh so efrain can we have these text files as well please after the session um look we could probably make the text files available for you and happily to do so um but i'm not necessarily sure what the text files will actually do without the actual topology and understanding the architecture so if it's something you desire a friend yeah i'm sure we can do it happy to help you help you set that up um maybe you want to do that and then kind of use that along with this i think that could be completely appropriate if that's something you desire we can we can kind of make that happen for you so basically i think the pasting pasting the config file just to save the time perspective and you can build your own lab and i think the concept is just trying to understand how the network configure and how you you do the the router what are the things you need to look at i think that's more important you can change any ip addressing you can build your own laps uh that's very easy for the to console so i think what we can conceivably do ifram i'll have chris or my team investigated chris is a whiz and figuring things out can we take these text configs can we take a picture or a screenshot of something like this can we place it on a cloud drive and share it in the description of this video for you to go and do it would that be helpful to frame because i want to make sure we give you the tools to be successful and if anybody has any questions while imran is doing some configurations please feel free to ask he's doing the build girl for the bill but any questions you know i'm here so anyway i can help along the way while mrem is building us a system so now on an r5 as you see in an r5 so we'll configure the the bgp ebgp between r5 and r3 and let me show you the conflict before i paste anything basically on r5 uh if you see the es number when you when you configure bgp you need to define the es number so r5 is like uh r5 is like router bgpas number you have to define the as number and bgp is it like a unicast protocol so if you won't appear uh if you want to appear with your neighbor you need to define the neighbor so if you see that the neighbor is neighbor 192.1.35.3 that is the neighbor for r3 35. uh 5 is the r5 so you're you're putting a neighbor statement telling them fear with this so bgp doesn't work without the neighbor relationship the other thing is you're advertising uh the new packs we created too so let me show you ip route so if you see um if you're seeing the on the routing table what exactly we are describing and configuring so we are advertising if you see so i peel out we are advertising the the the five network what what's seeing on the routing table and we are advertising the 55 55 network so the reason is um we are we are putting this 5.000 without mass because it is a default uh default uh network is direct is connected over here and uh the reason we are putting this from the mask is a subnet not the network here so okay so realistically speaking you know that's that that's the subnet so i want you guys to all look at this notice with every route you're gonna see something it's gonna identify it c means directly connected and me l means like basically local to the router or loopback okay so so that's assume we're going to have routing we're going to be seeing things from bgp maybe ospf we're going to show you how those routes are different and then at some point once we've got it up we'll just start adding interfaces and advertising and we'll show you how this routing propagates through the network that's where the magic begins yeah so we'll paste this conference on our files what are you typing in next um so bgp sessions yeah awesome so the one one of these when you do it i'd like you to do a debug ipbgp so people can see the finite state machine come up open xor that kind of thing not for this one first one but just one of these i wanted you to do that yeah r5 and then if you go for because it's not configured yet uh we have another neighbor r3 so if you go to the r3 you can hear me regularly mike you can see me that already i can see everything you're doing i'm looking to sell those interfaces come up i saw you configure bgp on the other one let me explain what you see here um the neighbor is the person on the far end you're going to put that you establish the tcp connection with and the remote ais is our autonomous system the network is basically something that all encompasses what you're going to put there and so on the other side on r3 the as number is 200 and the neighbor will be the 500 if you see the compare the config and the neighboring ip is 35.5 because there which is which is here 35.5 and it's all opposite vice versa okay so everyone i want you to really see this the top one says router bgp 500 that's an autonomous system router bgp 500. the bottom router says router bgp 200 no they are different numbers because they are different numbers this is called ebgp or external or exterior bgp note if they both said router bgp 500 it would be a bgp speaker on autonomous system 500 connecting to another bgp speaker and autonomous 500 that would be ib gp or e or internal bgp so when autonomous system numbers are different external bgp when they're the same that's all in the same family it's internal bgp back to your effort yeah so one very important point i just wanted to reiterate here mike uh what exactly we are doing for this network statement we are advertising the network we are injecting the routes into the bgb so there are two ways to do the injecting the routes either you advertise the network or you redistribute from somewhere else so just remember but there is a big difference with when you're advertising the network in igp versus a bgp because when you're advertising network in igp with the network statement it means you are advertising the local network which is connected network but in igp network can be anywhere so just remember it's very important concept the network command difference between i igp and bgp so here network is advertised remotely it can be other network but what what what is it what you're advertising what you're seeing on your routing table okay so i want to really re-emphasize and hit that part hard when you're typing a network statement into bgp you are saying bgp give this network information to somebody else when you're putting a network statement into ospf what you're really doing is you're telling which interfaces you want to participate in ospf and and form a neighbor relationship like hello are you there are you there are you there and the firearm says yes i'm there and then they start establishing neighbor adjacencies and start exchanging routes or llc's based on their protocol this is different this is you're saying i have this network somewhere in my routing table and send it to somebody else even if the network is not connected to me it's in my routing table with the igp it's enabling that routing protocol to talk on a certain interface like an arm to other people that it's directly connected to so that's kind of the thing so imran back to you absolutely thank you mike that's the video thank you really and you're doing a great job of showing people how to use this i'm so grateful now we have uh uh go show both bgp summary you see we have a neighbor relationship but uh so i'm on r3 and i'm seeing a neighbor relation between 35.5 uh what as is very important this is a attribute uh very important bgp attributes use a lot of functions it's uh the the two key functions of uh es attribute is the loop avoidance is the main thing uh [Music] it still helps to have a loop awareness from some pg perspective so if you see i'm not seeing any prefix even it takes a while what's the zero means we have a neighbor's relationship it's up and running and working but you're not seeing any prefix what you advertise the loopbacks so it takes a while to come back so if you see now now same thing you're seeing two prefix six so you're learning two routes uh from the r5 show here on r5 you're seeing the two two two uh two prefixes and your neighbor were as200 and another command to verify uh do true ip pgp you see you're learning the three network uh what is the next stop will be 35.5 35.53 is your directly connected interface uh we'll talk about later what path is coming from learning from ps200 so what is this 5.5 5.00 means it is a local route that's the loop one of the loop pack we created um [Music] interface brief see it's one of the loopback zero and we're seeing is next up is zero zeros it means local route the other look back no one thing no the route that you see in the routing table the bgp table is 5.0.0.0 note the loopback that he used to have that subnet on the router is 5.5.5.5 which is most likely a slash 32. so he's advertising something because he's got a route in his table that reminds you so with bgp it's not necessary to necessarily advertise the most specific subnets you need you may just want to aggregate or advertise your aggregate or your site or range or a summary route and that's what mram did here so i just wanted everybody to understand what's going on back to you emily thank you so run let me show you oh this is the new linux version so you see what what are the commands we are advertising or just might describe uh we have we don't have uh we are advertising the the network we are rising the two two packs and then the neighbor and soon when we get this all configured and all built we'll actually have some fun we'll install a loopback and we're gonna show you how long it'll take to propagate across autonomous systems which is why we'll say that bgp is designed for scalability slow it down ask it to do thus less computational overhead scales to more routes you couldn't possibly put three-quarters of a million routes into ospf but you can easily take in three-quarters million rows from 10 different in our service products in bgp no big deal now notice how the routes learned yeah so let's walk you to two things over here note you've got the 3.0.0 which coincidentally is an old-fashioned class a address you can also see that the 33 um subnet is actually the slash 24 which has been advertised note that you see a b next to them and why is it a b because it's learned via bgp note there's something this thing over here which says 20. cisco made up this concept of administrative distance what is administrative distance it's how believable it is bgp is a highly believable routing protocol externally so it's got a 20 the higher administrative distance the less believable the lower the administrative this is the more deliverable it's kind of like a friend if i call imran and say amram i had a network problem in my house can you come over and fix it i'd be totally comfortable about giving imran the keys basically saying here's the passwords of my systems go fix it and you know what because i know how capable he is i know that i could go to the beach teach a yoga class on the beach maybe finish the yoga class have lunch come back and and it's good well btp is so reliable cisco has said to give an administrative distance of 20 meaning if you get the information via ebgp believe it now ibgp has a lot more complexities and it's got a much higher administrative distance it's 200 right imran yep for ib gp meaning cisco thinks it's 10 times less believable back to you amram i just want to kind of try and walk people through the things so there's a few few important sorry a few important commands verification commands uh just for the audience there are there are a number of commands but a few important command i just want to type it here uh through ipbgb summary and so so show ipbgb summary and show ipbgp and of course i run this show ip bgp show ip route show ip or vgp so that's two verifications i wanted so now if you're seeing uh we have completed this ebgp session between these two so we'll repeat this okay on r4 let me remember one second we've got some new members subscribe and when they just popped on the call they said i'm not really sure i understand what's going on here so i just wanted to walk them through the topology that if you show that map one more time real quick i'm trying not to get in the way of the workflow as200 for the person that just subscribed and asked questions that is our data center as100 which we're going to connect to at a certain point is the cloud and we've got two direct connections to the cloud as500 up top is an internet service provider our data center is going to connect to an as600 on the right is going to be a second cloud say the left side of it as100 is going to be aws say the right side of it as600 is going to be azure we'll go back to your mram but then there's periodically people jump in and they have questions so i'm going to try and answer them okay so we finish this r5 and r3 we'll go to the yeah they gotta make things stop looking like each other icon-wise so we don't have to keep plugging the wrong ones yeah i don't want to paste uh something else so i do it all the time yeah so now we go to the uh we go to the multi-hop kind of a thing yeah so the uh the other thing we can do we'll configure this later on we can have a authentication as well yeah great yeah so we already have a neighbor so if we go to the r4 and r5 this is r5 and r3 so [Music] for the pgp 600. [Music] one second mike some from feedback from the audience sujit one of our favorite cloud architects that we know is texting me to say how much of a great job you're doing amrat thank you so i'm not normally picking up text but you know when it comes from our internal team members and others in the community so we'll go to the multihop we'll configure this r1 uh right now we'll configure the let's go to r1 [Music] we have the basic config so just want to add here a point um [Music] basically if you see on the on this test allowing uh we have two connection here with uh with ppgp and of course the two different systems different as a 100 and as 200 when when you configure the neighbor relationship uh uh by default uh by by default it is not uh it is it is not multi-hop so you need to enable mrt hop basically in order to have uh this connection basically ebgp always uh yd4 comes with a directly connected interface and here we'll configure this uh ebgp neighbor relationship between using the the low back but first first of all in order to reach the the network uh an r1 you need to define uh in order to reach the loopback in order to reach the the network you need to have a static route some kind of a reachability and then it's going to be worked that way so let me configure a route so i'm confiding to static routes from the r1 to reach the reachability of the the loopback so 11.7.2 and 11.11.2 is the e01 yeah and uh to note what he actually did in these routes ip route tells you that the destination that you're going to with the subnet master you're going to and then he put the next top ip address so that's what was going on there yeah ethernet zero zero is 12.1 xm0 12.1 and ethernet is 11.1 yeah so i i configure um i configured to 2.2 uh static route to reach to the next half 7.2 which is my which is my too much which is my the network here which is my network here on r2 uh any question mike um i think you know you're doing a really great describing this i want to point out one other thing so when he did the show ip is brief notice you have ethernet 0 0 that's up and it says manual notice that it says ethernet 0 2 and ethernet 0 note that it says administratively down when you look at these things they are shut down and if you want them to be unshut these interfaces as you noticed imran did a no shot so administratively down means it's manually down because the person didn't turn it on okay so now we configure the the bgp configuration uh typically you learn so this is the router bgp 100 and you're configuring this 198 168 2.2 which is a which is the 2.2 is a 2.2 is is a loopback it's a loopback here so you are configuring this as100 uh to a via the loopback and as i said uh the bgp always enable for directly connected interface this does not allow a multi-hop kind of a concept so in order to enable that we need to add another command by having a neighbor specifying the neighbor and by adding evgp multi-hop so mike what do you think is the configuration is complete i don't think so um yeah i still have one more thing because we are configuring the neighbor and neighbor is reachable by static route and it is not the next stop this is a multi-hop uh okay smart and other thing is because we are configuring is by a loopback so we need to have another command neighbor 192.168.2.2 update source loopback seeing the command mic yeah yeah i totally see what you've done okay so because the reason is the difference from this ebgp to this cbgp this is i this is not directly connected interface uh this is uh this is where we're doing this loopback the reason we do this uh your network is not always directly connected you might have a firewall you might go through the firewall so the reachability uh as long as your loopback is reachable uh the perfect use case now but typically we have a firewall somewhere in the between two areas so so the reachability is there so you can figure based on the loopback and based on the loopback you need to have a few extra commands telling me gp evgp have a multi-hop and update source look back to what i did here perfect yeah normally we just do evgp directly over the connection and we're not that far we don't have to do any update source loopback it's just neighbor on the far end of the ipa connection but you know and also you have to make sure that that's reachable because what if you noticed he added these static routes why do you have these static routes if the next hop becomes unreachable because you don't have a route to the next top the route may be in the bgp table but will not be placed in the routing table so that's why you got to put a next top static route to the loopback addresses for the ebs you can multi-out yeah that's for the reachability reachability perspective so so we'll do the same thing on the r2 any question mike from our audience um let's ask the question everybody out there on the audience we're doing some heavy-duty networking we've configured four or five routers we've done some static routes we've done some bgp configuration we're having fun with this stuff so if you guys have questions please answer them we'd love to take your questions all we want to do is help out so if you've got questions please bring them in i know this stuff is kind of complicated i've already spoken to mram and we're going to build this step by step so this week we're building a base thing we may come back in a week later and teach you some fancy traffic engineering or some route reflectors or two weeks later so if you want more network stuff type network training in the chat box and we're going to do it for you and you know if you've got any questions you'd like us to clarify let us know we want to be your source for information we want to give you this information we're going to help you get to your goals so let's get to your goals libra um fantastic we love hearing those um can't wait till you start and i gotta tell you this networking stuff is really great dram you're right a loopback address is virtual and it will never go down let us know by typing network if you're loving the networking stuff we will produce so much more networking content and we'll get you there nick love you said networking library use of networking i got to tell you all i am alonzo networking i am having so much fun doing these kind of live training sessions online reginald jam wonderful i literally spent as much as a car would cost to buy some new special streaming gear and it's going to be here in the next few days we are going to have daily show we're going to talk about cloud things networking things what's going on in the community we want to be really hard to give you guys some extraordinarily good free training and for those of you that want to participate in our pay training programs i promise you it'll be a life-changing experience for you because what we put and then what you would consider to be a days pay for the cloud architect somebody else would charge 20 25 000. we teach executive communication presentation skills architecture skills the network the data center everything so you want to work with us directly we love to work with you you want to reach out to us and use some of our free services we're excited if you want to do both that's great too all we care is that we help you get to your career goals so ask your queen training i see you guys a lot of people are liking the network and guess what we're going to be doing a lot more networking training and i'm going to make sure to bring my expert and good friend imran with me on a lot of this stuff so make sure you know it from the best and the brightest people in the world that we can find so so mike you're seeing an r1 192 168 2.2 which is my the loopback here i'm seeing two two routes uh pcmp equal cost load balancing so the reason we might have a similar case all the time we have two links and um i'm seeing two routes this route is learning from 12.2 11.2 do you want to add something here yeah okay so let's look at that it says s and what does s mean it means it's a static route note that one the administrative distance of one you manually told that router hey to get to this route here's your next top it's pretty darn believable because you manually told it that so it's got an administrative distance of one which is about as low as it gets so that means reliable reliable reliable note he put two equal cost static routes in and that's why you see two routes to the 192 168 2.2 32 subnet because he has two equal cost routes if he had wanted a lower cost and a higher cost and he can change that if you wanted to change the metrics the root you could see immediately it would change that but two equal cost therefore that you're seeing what you're saying okay now what is this showing explain that remember yeah so i'm i'm learning you know i'm sitting on r1 i'm learning a two network uh and uh the next stop will be 2.2 and so we're learning two networks and we're learning a spot 200. now r1 which where's the location of r1 so we can answer muhammad's question yes r1 is here so r1 is there we're not going to be running osp we would be we will be running ospf to the multiple routers in that area for network player reachability absolutely we're also going to be running ospf and all the in local areas for network layer originally so we will be running ospf from r7 to r1 that's area 0 and then we'll show them the redistribution of the redistribution work by a network statement and by redistribution ospf aspf area and fantastic and after we build it i'm going to have you show people if we change the weight on one side prepend the another route change the local preference set a community on one side match that community with a route map and change the the way we'll play with it we'll show some tuning capabilities but first we've got to build it hope we answered your question there muhammad is it good it's great okay so now now go to the see we see we're having r7 and r1 we're running the ospf we will run the ospf here there's nothing on r7 just add it now without the hosting so we'll configure it at zero zero and use the 17 network okay so let's everybody see where r7 is real quick just to make sure so note r7 is that is in that that's we're gonna be our cloud vpc um and uh for example r7 will be running ospf to r1 for intra that internal as 100 network layer reachability information so i'll configure the e0 [Music] check i configure loopback 0 and go back one you see i i define the isp ipo spf network point to point yes explain to them why you did that uh [Music] typically speaking ospf generally speaking is a broadcast multi-access it sends an ip multicast hello to 224.0.0.5 it says are you there like a health check are you there are you there and it listens to a hello from somebody else and when they form it they form what's called the neighbor adjacency and after the neighbor adjacency is formed they say send lsas now if you're on a broadcast multi-access network like ethernet you have what's called a designated router which basically is the primary model for that subnet and a backup designated router and those two for what it's worth listen on 224.0.0.6 for multicast and lsas and things like that so point to point is because basically you're sending it over a point-to-point link and because of that you're basically saying how you want it to form any of your adjacency instead of just a multicast alone like you would use in a broadcast multi-access network yeah thank you jeannie i love to see the enthusiasm on networking gina i love all the enthusiasm one of those wonderful things that i see when we interact on a frequent basis so i'm configuring the ospf from advertising the oh okay so what he's really doing here with ospf when he's putting in this network statement that 192.1.1.170.0.0.255 he's basically saying add all interfaces that match 192.1.17 and anything else in the last talk it up bring those interfaces into area 0. turn on ospf start saying hello hello hello start establishing relationships and then after those interfaces start there they will start advertising their links ospf is a link state protocol basically lsas are flooded and everybody sees the state of all links then the routers use it's his or her algorithm and they run a short as far as path calculation and then they say ah to amram that way to chris that way to genie that way and that's because they're coming on the network because i'm turning them on on the interfaces and those interfaces know to participate and those interfaces know how to advertise their links perfect cameron so i'll go to the r1 we have to do the r1 the same content so there are two networks we are advertising uh one is the directly connected ethernet and the low pack so i'm configuring e02 uh which is a directly connected database or ospf so that's the link between the two yeah r1 and r7 i'm just putting the ipad address and no shut and then of course i add the ospf uh see the links are up and i can put the usb of config area 0. now you're going to see the fun on r1 if r2 is configured with ospf and r1 is configured with ospf guess what we're going to start learning some rousey ospf soon no okay so what do we see here we can see that the router is full and it's also the designated router so we know the neighbor is in charge on this section it's the designated router which means we could probably be the backup designated router because we're the only other router on that subnet but we're not sure yet and that's the designated router we will be listening on 224.0.0.6 for lsa advertisements even from them even though lspf routers are 224.0.0.5. so amron do you want to show what the routing table looks like well first let's show people where r1 is in case they forgot in case anybody is like me with a short attention span if you can show them where router 1 is in the topology and which routes it's going to be and it's going to be learning routes from router 7 over that ethernet 0-2 so let's show people what the is what the lsa database looks like let's show people the routing table let's walk them through it okay on route we're on router one right router seven right now yeah there's no learned routes via ospf yet yeah but on router one john what so we see we've got a full neighbor so let's go figure out why of course we may have just advertised our own routes in here yep see okay here we go now look at that routing table okay we've got some c for connected routes we've got something with an o next to it and it's because it says an o we know that it's inside of our area if it was from another lspf area for people that are not used to lspf what we do is we divide into areas to segregate our network we have this area 0 which is the concept of a backbone this is where we put our highest performance routers see what goes on is everybody inside of this area will know about every single link inside of this era and then we'll run what's called an spf calculation on it that gets computationally expensive it really runs up the use of the cpus so from an architecture perspective you kind of decouple things you modularize things guess what all those decoupling and killing things and micro services and all that stuff you heard about now this is 30 year old networking for people like kim ramona what did we do we took a network we ran ospf a link state routing protocol and guess what there were too many computations in our network so we decoupled it and we segmented it so we created the concept of areas the ospf area zero the backbone knows everybody's everything it knows all of our stuff and we get summary information from external other areas and they're propagated very differently so by segregating our areas we can do things like create stub areas or totally stoma areas or mass stubby errors we can reduce the size of the database so you want to show what the osp show ip ospf database or lsa database or whatever it is okay so here's what you see you can see the series of links that you receive and you can see where you learned the links and you can see the age in um seconds so you get a sequence number which is message number you've got to check some what else you check some basically it's a one-way mathematical house to determine that your data is good so i just wanted you guys to see this is what goes on in between databases so um go back to that routing table if you don't mind and shapoor welcome zaki welcome genie welcome more networking yeah let us know in the networking window that you're having fun with networking while emma's busy configuring this and explaining it along the way okay so look at this this is what you're going to see in a railroading table connected route ospf learn routes bgp learn routes static learn routes now going back let's look at this this route is learned via bgp it's 2.0.0 eight note its administrative distance is 20. note this next route is learned via ospf 7.068 its administrative distance is 110. guess what if we had a route learned from evgp to 2.0.0.8 and we had a route learned via ospf of 2.0.08 which one do you guys think is going to be believable the one with the administrative distance or believability factor of 20 or the ospf one where the believability factor is 110 due to the administrative distance note the lower the administrative distance the more believable the router is so which do you think is more reliable an ospf learn route for a subnet with an admin distance of 20 or i mean a beach i'm sorry ebgp learn route with an administrative distance of 20 or an ospf learn route with an administrative distance of 110. pop the answer in into the comments section okay see ingram's getting a little fancier see when i started we just did a show running config we'd look at the configuration now with the newer versions we can pipe and grab and do all kinds of cool linux stuff and that's what he's taking advantage of here see see on r1 we are not uh we're receiving the uh the ospf route we are we are we are not redistributing it because the way to redistribute an r1 the way to redistribute r1 or spf there are two ways to do it either we redistribute it or other we advertise a network statement so firstly first we do the network statement then i will show you the the the difference uh different of those routes and then we will remove it those routes from the network statement because uh like so [Music] let me put the advertised by a network statement so i'm i'm advertising uh the the seven network and 77 network uh by using the network statement and a bgp on an r1 okay so we've got r1 we've got our neighbor adjacencies up right we've got vgp configured r1 do we see all the routes we need to from our data center and the data center now see all the routes from um what do you call it from the uh cloud yeah yeah so okay good have we configured routing to the internet service provider next yeah so if you see like like joe ip do pgp when you see a bgb table these routes will be coming as an eye and seeing an eye here i hear heard i i means is preferred over uh ebgp apgp like a question mark so i i always prefer the evgp because the reason is uh we put this on as a network statement that's what is showing that so now the other way to do it of course you have 100 routes it's just like two rounds you advertise it by using the network statement but you have a 100 routes coming in uh you will not do this this way so you redistribute the whole ospf domain then you control the the policies what route you want to see so i will i will remove this uh just want to show that thank you for demonstrating that that was really good so i'll i'll go so i'll remove this from from the r1 i'll go to the router dcp 100 and i'll remove this so how how do we remove the network statement no no command and add it to the whatever the router just added we're learning from our routing protocol uh we are going to remove it and show you how to redistribute the other way the better way uh just understand the concept of networks returned and redistribute things it pretty it's pretty amazing that the no just reverses everything um best way to remove it yeah so if you're going to show ipvc [Music] you're not seeing those routes seven and seventy seven okay now you're you know you're you're going to redistribute this uh you're on the i'm on the routing protocol pgp domain and i'm gonna redistribute ospf everything is ospf one okay so what he's doing here now is he's taking the information learned via ospf and then placing it into bgp so bgp can share it to other people that's called route redistribution note you could put a network statement it could have you know redistributed connected he could redistribute ospf subnets and then of you and then and that's how you could do this so that's kind of the information because we've got to get our information into bgp to then push the information to someone else via bgp so so if you um if you see the the reason i do this because when you learn the route over there so it's now you're seeing the question mark you see that uh the the question mark is uh uh is incomplete for incomplete rounding information yes correct so you want to talk about the origin code over and how which is preferred yeah exactly exactly that front right now i want to show them how do we change this so you're learning that out from question mark in earlier if you see on the top you're learning the routes by i with the origin code so i always always prefer uh the uh the question mark want to add something here original i think it's you think it's great you're talking about the origin code you're showing how to change and i'm thrilled yeah so i'll put in an out map and change it uh so to i and there's a way to do it so that's what you're controlling routes on r1 okay route and apologies for some people as imran's doing this i'm typing commands back i have been a martial artist an olympic weightlifter for many decades and let me tell not of these days after i shot her hand on foot i'm training but i will tell you we've got lots of arthritis so my temping is not so pretty so if i type quickly to try and respond to folks on this call apologies for the bad spelling so arthritic hands aren't too clean 100. so we again we are redistributing but we are redistributing with a route map [Music] right now and by doing this by redistributing with the route map he can change the origin code to make something more preferable for example you want to explain how you know the preferred is you know generated in one way versus another way there you go so the route map he can name it anything he wants and here is basically doing something that's going to match something else and just change the origin code and when we're done this we're gonna we'll demonstrate you by walking through the routing tables from end to end and guess what we might use a route map to filter something or use a community or something like this will get us out a little bit of fun with the bgp attributes networking is cool at least for me it's always fun to watch the way these things work [Music] and now he's got his bgp configuration you can see he's connected to the external neighbor and remote as200 you can see he set up egp multihub because he's going loopback to loopback not exactly directly connected and he's told it to update from the loopback so now you can see what's going on here now you as you look it will no longer show incomplete you you want to show the ipbgp table yeah for some reason it's not earning um sometimes this should never happen but sometimes what's going on is when you're connecting to people in bgp and you modify something some data gets corrupted and you've got to basically re-establish the session kind of like your cache that gets corrupted i'm not sure that's what's going over here but no we should test it let's do um we set the origin code but did we match anything i didn't see any matches in the route like i didn't see like a matching access list i didn't see a match but i think maybe the names are incorrect and this is why you build your configurations generally speaking at a time because this kind of will happen yeah let me remove this but we need to match something either matching the ip address of the interface create an access those match the letters but it should change um just redistribute uh yeah it's coming in see the the route map names incorrect so that's why it's not catching it you know the type of yeah typos will get us every single time yeah now now you're seeing uh as usual change it to distribute now you have an ospf route with the i or it report earlier you're seeing a income incomplete thing so does that make sense now mike completely so you can see where the routes are actually coming from so now which we're on router one so go but going back to the topology you know we see if you go back to the show ip route on router one um just so we can go back walk through what we see okay so let's look at that whole routing table okay so here's what we want to see we want to see who's learning what from where so we're learning the seven knot subnet from ospf inside of our area the ebgp we're learning which means it's coming from our personal data center we're larger than 2.0.08 via ebgp we're also learning the ebgp the 22.0.0.24 i'm sorry so for all those slack messages coming into imran's desktop um the reason when we talk about being a network architect or a cloud architect is being really brief and to the point you can see messages directed to me or popping in on mram's laptop for people submitting homework assignments which i love to grade so because that means my students are learning so just realize when you speak to executives the reason they have such a short attention span is that thing that's going to amram's desktop that's related to me doing things that's just one of the many millions of email messages i get during it's not millions but it's a large number so realize executives are that way that's when you hit an executive fast to the point before they change their mind and their mind goes and drifts to some place off so good routing table we get to see this no there is no default route so if we don't have a route to somewhere else we are just going to drop the packet and that's the way it should be we don't need a default route right now because we're running bgp to an ospf or everywhere so perfect now um we have our data center do we have we have bgp up and running in our data center or do we have we have just ebgp or did we set up ibgp and we'll set up my bgp as well uh mike between r2 and r3 uh we'll set up a ibgp and we have a we have one on on like r2 we'll have a set up a loopback and we we set it up a like here we set it up a ospf where we set it up eigrp between the reachability perspective the one i set it up and we see how that goes uh uh with ibtp excellent so why don't we do this can we come to router two and show what we're learning from router one and then hop on over to router three and show all the stuff that we don't know yet without ibgp before we move over to things [Music] and of course while we do this if any of you guys have any questions or any desires to ask or learn something tell us if you guys want to learn something if it's possible i'll make content and generate it for you so tell us we really really want to do what we kind of help out okay looking at this look at the show ipbgp look what we've learned we've learned about two network meaning it's originally generated by us the 7.1 we learned about it from the next top you can see it the 77 we learned about it and the network can you do a show ip route 2 so we can actually look at the entire routing table okay so note we've got some b's and c's what do we not have yet we have no ospf learn route from our internal network and if he goes now from router 2 to the router on the top of that triangle in our cloud please note what routes actually gonna exist in there and then can you show that routing table oh where to where right sorry so okay so i need to see the picture to know which one yeah if you can just slide the configuration down so i can see the topology real quick this is r2 here we are so we can go to r3 or rfr and show that none of the uh for example none of the routes via bgp are learned there because there's no ipg oh you have you have ib gp running sonar 3 okay so the bgp configuration here okay so okay the routes that are being learned from this guy are being learned from the internet service provider at the top right no there are no routes learned from r1 here then no no so the routing information on r1 is not the same routing information that's on r3 correct we're not we're not learning anything from the ospf here okay so let's stop for a couple of seconds before we do there's a couple of questions from the audience i want to draw someone doesn't understand what a route net works okay let's be fair the route map is coming cisco certified internet expert level work what a route map is as follows if i want to take in a route from something let's say amram and i are ebgp neighbors i'm communicating with emra bgp could send me something called imram could send me basically a community called community red i could then take in community red and community red is arbitrary it means nothing to me but i could then take in community around and say match community red increase the weight which would prefer increase the local preference lower the med so the route map is really some advanced cisco router configuration that enables you to take something and do something else with it the route map works almost like the code the programming language basic so imram says me let's say he sends me community red i could say match community red and increase the weight of this subnet so i hope that kind of made sense a route map is just a means to tune something the singy um we will have another session for follow-up and you can ask any questions you want anytime to be thera basingy this is real network training we have gone in the beginning to ccma level then we've completely exceeded ccnp level and we're kind of hovering here in that ccnp ccie level i got to tell you if you want to be a cloud architect on the network you got to know this kind of information so this is kind of the basic networking and this is heavy duty networking this is serious networking because when i say if you want to be a cloud architect learn the network in the data center this is the network and you got to understand this and when you go to the cloud and they talk about bb bgp or they start talking about cloud hub or transit gateway you're going to see that this is all just this stuff so don't worry i've spent 20 some years studying this amram spent 20 some years studying and mastering this i don't expect it to all be there initially you can watch this video multiple times we did a completely free aws advanced networking which is kind of like an intro to networking the link is in the description below but my my associate chris will pop that there and don't worry you will learn it step by step by step by step we'll break it down into what you need we're going to do this again and again and again and don't worry we'll get you there here's the thing the singing first you need ipconnectivity what is connectivity um basically you need an ip address and you need a link and that's connectivity then you need a path to destination how do you get a path to the destination you either put static routes everywhere or you run a routing protocol that's what we're doing today we're working in routing programs please understand it has taken me year after year after year to learn these routing protocols so yes kind of look at it amram has spent years mastering these routing protocols i spent 10 000 hours on just bgp so when you see people that's teaching a networking course like if there's somebody a certification provider making a networking course that's never worked in the networking in their life and they advertise their events network of course every day realize this is just internet and networking even here we're getting in this ccmp ccie area so if we have to do this a few times that's great if we have to educate the providers that are prevented making courses out there about networking we're happy to do that too all we care is about the club community so so just to add here the reason we start took the bgp because we want to map this cloud networking piece into real networking we need to learn the bgp very well so most of the routing protocols on the cloud site you're having a pairing of a single connection dual connection uh and how you manipulate your the bgp route with different use cases so just this is just a simulation and we have more to come and we can start this this is too difficult we can have start from basic to to a higher end but this is really a complex scenario we just want to show you how the routing works the concept wise and of course how you manipulate routes by using different pgp is therefore last more than 20 years so so as mike said there's there is there is change uh everything is in the cloud but the routing protocol is the same and the same behavior so we are just simulating different clouds here data center and uh you have different fears and you can you can manipulate you can return this topology or anything else right so i was looking at peter's question peter i hope i was able to answer your question about the route mount the route map is this i have information i'm going to take that information match it and then do something with it so i could create a million to one route mods a route map could be if information comes in on my left arm send it out my right arm or a route map can be match this do something so really all the route map is is just matching something and we probably went a little too deep with the route map really first um but realistically seeing um you know we'll we'll show that we're going to build the basic nature without the route map first get all the routing up and then we'll do this stuff the route match is the tuning factor roadmap is just access list controlling your routes what's need to be there basically that we call it traffic engineering what's need to be inserted what needs to be going for this specific uh attribute for bgp or it can be anything you want to allow any specific network and you're having to allow deny any specific network so we're controlling the traffic uh the route map is a tool uh you can use it in any other routing routing protocol so what our car do now uh between r2 and r3 we i create a loop pack uh and create a because we have a ospf here we have it uh we will use an eigrp here and see how things work in the agrp and then we see the the reachability between the by using the the low pack so in my r2 let me go okay excellent so so i'm i'm creating an r2 as a loopback sounds good we're gonna set up set all this up get everything up and running um peter hunt basically exactly if you match something take taking action on it that's exactly what a route map is excellent excellent sorry imran white you were doing such a great job with the networking i was trying to make sure i addressed something i i just created just create a 10 on r3 and r2 and we run eigrp between r2 and r3 and what we advertise this network this connected network once 192.168.23 network and uh of course we'll advertise uh this this network as well so i'm on r2 so on an igp mike we are having uh two networks we are advertising they are directly connected this one and this one and this one and this one and we are configuring uh eigrp so okay so note the loopback is up yeah so you're running eigrp inside of this autonomous system that's somewhere okay so for fun we've chosen to use another routing protocol as our igp in our internal area lots of organizations use um eigrp because it's very simple to maintain lots of organizations use ospf big service providers use intermediate systems to intermediate systems as another type of link state routing protocol so what imran is going to show you prior to setting up the ibgp is when he configures these three routers in our area note they're all going to have different routes they're all going to learn each other's routes for these three routers did you set up eigrp on all three i'm just setting up i just want to show the audience this is the commands i'm pasting it grp200 this is the network on the one side 23 and 24 and as i said when you use the network command in in and igp protocol these are the network uh locally connected so and i'm advertising you see on the just paste it over there so if if you see the an r2 that's 23 and 24 network and this is like directly connected interface and and of course the the loopback 10 which i created so all these are directly connected network and um i'm reaching this by uh putting this on on a year grp mike you're good i am totally good so i'll do the same thing router eigrp 200 and 34 23 and 34. and paste it see you can see the agrp neighbors are up between r3 r2 and i'm going to put it r4 so and muhammad exactly you got it um just if this match this so imran is going to set this up um for those that are on the call if you could leave a like or share this message or invite someone if you find that you're getting some good training we always like to share our content with as many people as possible so i create a look back then create a eigrp on r4 as well yes so we don't really use any reserve addresses here i'm not really sure what you mean that when you do it with networking outside of the cloud basically the subnet address is reserved because you can't use it it's used by the router for the route on the network and you can't use the broadcast but otherwise all addresses are good on the cloud it's a little different like the first five addresses of the subnet or four addresses of the subnet are reserved for experimental purposes and such and then the last one is the broadcast so it's a little different here okay so look what's going on here so if i i'm on r2 when i do show ipa grp neighbor you're seeing um i'm learning uh i have a two neighbors okay so looking at this right now he's got two neighbors so he's learning here yeah he's learning ergrp routes from two neighbors so regarding auto summary well there used to be this concept in eb and ar grp where it would naturally just summarize routes the classical borders which was an awful nightmare for all of us so we all typed no auto dash summary it's probably we're using ios version 15. something do we have to disable no auto summary in this or is it disabled by default so uh if you want to verify uh of course we have two neighbors here and our poor and see so now as i said uh we have a reachability uh it's an r2 so can we bring the r [Music] r3 to back address so what he's going to now do is test the networking to make sure he can see it see i can see the the low pack uh rod three so he sent five icmp echo packets and he got five icmp echo replies back so now look at this okay what do you see in the routing table you see you're directly connected you see routes that you're learning from bgp and you see these routes that have a d next to them the d is coming from e i g r p internally learned route note he's got two equal cost links and because he has two equal cost links you can see them as 90 slash the actual link cost you can go from here you can go from there so andre asked a great question is there a great way to learn the configuration language well let's put in the concept when you know what you want to do it's easy to look up the commands on the cci exam um they basically ask you to design something that's super complicated everyone's done this with two cci's he's doing the third one just to prove his son hey by the way i can do it which makes it my hero i've done mine back a million and one years ago and it was this miserable two-day test and loved it um but it was fun um and they would literally hand you the cd with all the configuration commands they just want to know if you know how to do it so learning these commands um typically speaking many of these commands are learned at the ccna level the rest of the commands are really learned at the ccnp level cisco and their website has all the free configuration training you need knowledge of knowing what to do and how to do and how to design and how though that's what we're doing today that's not free um that comes with training so configuration bands always easy to do cli is easy when you know what you're trying to do so just by working on this you learn it okay so now we have this working we have a reachability for the loopback here by igp so now as we said we have to configure ibgp because whatever we with the route will learn from as100 is need to reach on the both edges mike is very important yeah perfect so whatever we learned from this is the as100 or a600 just need to reach uh from here to the uh the exit of the edge with the right property so that's why we use ibgp which is a having a support function for epgp i just want to re-emphasize the code perfect so i want to make this clear let's out ibgp running here show that there are different routing tables on r2 and r4 can you show that so the full routing table here i want everybody to really see this okay so this is the you're on router 2. you're getting some bgp learn routes from the top internet service provider correct and you're getting some eigrp allows from the community but you've not configured ibgp right you know right because i was answering questions so note on router 2 you're not going to get any routes from as100 where the cloud is yet you won't have them on this router because realistically speaking you uh haven't learned them yet so we're going to need to run ibgp across our network in order to get these routes we're going to try and avoid things like redistributing ospf and eigrp once they make you do it on a cci exam there are times when you need to do it but no matter how you do this no matter how you set up your prefix listings and tribute list this is right for actually causing routing loops but running ibgp across your backbone will pull it all in and enable you to connect it to other autonomous systems so what's gonna happen is imran's gonna actually set that up now he's going to show you the routes on router 3. see you're not here so you're not you're not learning anything any bgp routes here good so you want to show the ip routing table from this one okay so we're learning two routes from bgp a little bit note we have a lot less routes in the routing table on this one why we're not pulling them in from the cloud provider so what do we've got to do to make that work how are you going to make those cloud provider routes be available to all the other devices in our system yeah so we need to have ibgp working and between the between these two and these two and these two so it's like at this moment we just have ib gp we have a reachability and now we are we will configure an r2 r3 uh as an ipgp protocol uh with of course slightly difference with the neighboring command and upload source loopback and the next hop next stop as well so let's set up that ibgp and when we have this network configured and it's fully finished we're going to get to have the fun we'll do some trace routes we'll start manipulating traffic engineering policies but we've got to build it so but this is the infrastructure fundpar if you you got it type network in the chat box i know you're not sleeping please let me know by typing in network or yeah network okay so i'm configuring ibgp okay so now it's going to set up the ibgp so as he does this let's note what he's actually doing when he shows you the configuration so if you see the the config here so i'm i'm putting as a router bgp as my local ais and i'm neighboring with standard 3.3 which is a remote on the top r5 and this is the loopback but we created so we are doing a neighbor relationship for ipgp as a loopback and we have we check the the standard 3.3 have a reachability and we're putting in a next update source will be the uh will be the loopback 10. and further it's very important command uh uh we're adding this next top self uh command we're telling uh uh telling the neighbor uh uh next top cell and how to describe it in excel fields are very important okay so here's what's going to happen if the next topic is not reachable the route will be in the vgp table but the now it won't be in the routing table so you can't reach it so what's going on because we're using loopbacks and we're not actually using that physical link where you would just be on the link we kind of have to give it the next top which is our self and because if the next top root isn't our routing table we need to make it sort of there somehow or else the route will be placed in the routing table so that's what he's doing no with the neighbor statement the remote as is the oops didn't mean to do that with my mac screen but the remote as is the same number as our autonomous system and that's why it's ibgp note if that remote as was 300 not 200 it would be a different organization so when you connect to the cloud it's always going to be a different number than you want and you connect to you it's going to be you same thing r2 and r3 they are on the same yes you can notice out of pgp see no okay there you go now in router three you want to show everybody where router 3 is on the map that's that's our over 3. okay so look we're actually can we slide that thing down a little more so okay so look at what router 3 is router 3 is now you know at the top of our as200 that's our data center so this is our data center now on router 3 if he's going to do a show ip route we should be able to see routes from router 7 sitting in the cloud and we should also be able to see routes that are coming from as500 because we have ebgp pairing there so on router 3 we should have a lot more routes than we have before so let's see what we got there emerald um r1 sorry i'm route 3 for the full routing table it's going to be a big routing table now okay so now now you're seeing uh this is coming from bgp routes here no now look at this now look at these routes so we're going to see all kinds of things we can see for example that we learn certain subnets via ibgp note that 200 that's administrative distance so if we have two routes in the routing table for bgp or we have two possible routes we get a 2.0.08 with ebgp and we're going to learn the 2.0.0 eight b e i jrp eigrp is administrative distance is 90. bgp's administrative distance for ibgp is 200 we would put the other one in the routing table now look at the routes we learned via ibgp note they have that administrative distance of 200. by comparison let's look at the route to the 5.0.0.8 subnet note we're learning that the ebgp to our upstream neighbor so different administrative distance different believable but what i want you to see here is now the routing information that we're learning from that isp up top we have inside of our system and all the information we got to the cloud we're still maintaining on our system so we've got a lot of stuff going on right now of course mram saving the configuration with the right mem and before you get to the next router i'm going to ask sir a very good question from mr andre he says in real life what are the steps design configuration customer the configuration steps the system and the way it starts for the architect is as follows mr andre start with the customer ask the customer what are their business goals what are their business challenges what are their pain points then the job of the architect is to go back and think how do i enable the customer's business to work properly probably that means what kind of solution can i design so with the architect we design that solution now after we design that solution somebody's got to go build it so we turn it as architects we're kind of done we turn it over to an engineering team to go build it and then when the engineering things don't build it they turn into an operations team to go maintain it so mr andre i hope i answered your question this is the role of the cloud architect with the network architect it's the same thing meet with clients figure out requirements design and then turn it over engineers to go build it so that's kind of what the steps are i'm sorry back to you oh that's fine that's fine so we have a ibgp relationship built up between r2 and r3 and now we need to configure between r4 and r3 which is if you're saying r4 r4 and r3 that that is the part so and when you go through these if you could just remind people each time where the routers are on the uh on the map because it's hard for all these routers yeah so this is r4 and uh this is router's r4 and we can figure this idp here and the same thing between our r4 and nr r3 basically and okay so you're gonna you're gonna set up uh router three you can see where it is note when this is here he's gonna be picking up routes via ib gpa then at some point he's going to connect it to that second cloud we'll call it azure so once we get all of our routing up and running that's when the fun begins that's when we can start tuning we can start teaching we can walk them through but we hope we're trying to make a good examples of what we're doing while we're building we hope you kind of get the feel for this we understand this is very complicated technology this may not be possible the first on the first time we produced a link from aws advanced networking content it's kind of like aws advanced networking is kind of like an intro to networking course because of that we made a completely free course chris from my team posted a link to it it is on our youtube channel but this is much deeper this is the kind of work an infrastructure architect would be thinking about or even a good cloud architect for that matter so on r4 we're having a router bgp and we are rotating uh by my uh 1033 that's a loopback r3 we're pasting it here and r3 we do the same thing so if you go to the art form type one nice so if you if you notice here we are learning the the 77 network uh as200 uh and yes 100 which is the oes 200 is here i'm i'm i'm i'm on r5 which is es500 so i'm showing you the the network sorry okay so now what you just see on r5 unless we run any filtering and we're going to have some filtering at some point what you'll notice is we're going to have routes learned from the data center and the cloud our data center and our cloud so on r5 he does a show ip route no you're learning all those subnets via ebgp so that means they're reachable and for those people that are not familiar with bgp chris from my team will do the following we've created several videos on bgp that are sitting here they're the ones critics that begin with like intro and networking and they have aws bgp with a pipe spaced in between them if you can link these chris if you can put a link to the bgp document that we wrote which explained all the bgp attributes and how to manipulate them how to optimize them please place them two chris into this window oh so the the the the thing is we're trying to show it like uh is having a right uh the right uh the right as because even though even though it's coming from [Music] igp it's coming to the right is coming from 100 and 200 uh that's the reason that's the reason we show you how to use redistribute it and the reason is ibgp again is helping helping ebgp to change do not change this attribute see see this attribute does not change and this you're seeing is taking the right as path to avoiding any loop no no provided so okay so really i want to just try and take that and run with it for a minute no this route 77 it's a s path remember when i talked about in these videos one of the things you can do prefer the path with the highest weight for further half path with a local highest local preference prefer the shortest as paths if you've got more hops over in here it's going to be a longer asp so later we're going to have some fun by tuning aspas but for now um i just want to elaborate let us describe the path that the autonomous systems your data is going to travel through to reach its destination okay what is yeah that's how much is happening okay well we see this what do we know we know we have one of two problems we either don't have a route to it do we have a route to this in the routing table and if the answer is yes the next problem is they probably don't have a route back to us but we can verify this and we can see where the traffic gets lost with the traceroute but the vgp manager that chance to converge either yeah so let's let's continue we can troubleshoot later but but as long as we understand the concept it will be it will be it will be very valid so so to do now i think uh i think the last part i have um is mike um before we end up with this thing is configuring is r3 or the rr rr before we do the route reflectors do we get all the bgp running to all external autonomous systems everything everything is configured i believe let's just check verify everything and then we're going to start having some fun yeah this is r5 i'm learning two three five two three a and five two three and five number five [Music] i know there's some complexity with some of this stuff but nothing worthwhile in life is ever super easy and i can tell you with persistence perseverance and a little bit of work you can learn this network which will put you leaps and leaps and leaps and bounds above everybody else in the cloud architecture and if you're trying to get your first cloud architect job or get promoted as a cloud architect or take on bigger roles realistically speaking this is something good for you to know now chinton don't you need routing between two and four um i actually have to see the picture to even remember where is two and four um two and four yeah two and four um without the picture i lose them out but with two and four i i'm assuming we have to have routing between them [Music] the the network is directly connected over 24 and you can see you have 24 yeah that's you know what while we're here okay the routing is fully configured while we're here you know as architects sometimes you meet with the client and you ask them what they've got and occasionally you know they don't know now pretty scary but occasionally someone just doesn't know what their network does so people like imran and i have literally mapped out networks by going from router to router to router by doing something called show cdp neighbors so if you want to show what that shows we can actually find who our routers are next to us through this layer protocol called cdp uh sis what's the cisco discovery protocol oh yes uh uh mike uh just need a one second break uh sure okay we'll come back to me i'll start asking some no answering some networking questions for a little while i am always fun and happy to answer some networking questions so while we wait for mram to come back and start doing some configuration stuff do we have any networking questions so far iep addresses routing protocols subnetting anything we need to cover while we're waiting for imran to come back in the meantime people can type cloud so i know you're there that's great and if you're having fun if you can leave a like or a subscribe please do so because that signals the algorithm to then share our content with more other people and we're going to produce as much free and highly valuable content for the club community community all over the world we realize that not everybody can afford our training and we want to make sure for those who can't we still provide ample quality training so if you need to take so if you guys have any questions um please uh let me know so while we deal with a few things you know internet connectivity things anybody have a question for me because we got a couple more minutes and then we'll go back to camera any questions on the network any questions on routing protocols alonzo you're having fun with networking that is fantastic regional cloud architect let me know you guys are here by typing cloud architect leo asked a very good question what is the advantage of using er jrp instead of ospf i got to tell you 100 years ago when i started networking or i should say 25 years ago er grp was used very frequently because it was very simple to configure ospf was a cpu computationally expensive protocol eigrp was light so organizations used er jrp because it was simple and elegant and a little more tunable in today's world for the most part i don't know many people that actually use eigrp for the most part what i typically do is focus on ospf typically speaking and bgp eigrp i don't know in your experience imram if you're still using it a lot but i stopped using it about a decade ago yeah i think thanks mike uh there are customers still using eigrp some of their flat tank words just to add here eigrp is a cisco proprietary and ospf is of course open standard there are certain uh customer use case small flat networks people love it very easy to manage so it depends on the scale size of the network customers still use it and they love it and they keep get using it depends how scale in larger networks now typically the larger effects of spf and isis and of course from the cloud era mostly in the bgps absolutely so next question from muhammad with real data centers you will define a routing protocol absolutely typically speaking in a data center you're going to use ospf all in a single area because it's a but and usually you may have to constrain it but typically it's ospf in a single area inside of that single area ospf everybody will know everything about everybody else's routes and you'll have very fast routing now outside of your data center when you connect to the internet you are going to be using ebgp to to put that into context but inside of your data center you will all everybody will be using an igp like this so the singy just to recap from the presentation what is as100 200 500 and 600 they are different companies and different organizations so one of the organizations of our specific environment is our data center one of the the information up top i believe that's as500 is an internet service provider on the left side of this document we've got an aws vpc and we're running ebgp to it and on the right side of this document we have an azure-based vpc and we're running to it so i hope i answered that question for you so going back to imran because now he's back to a position where he can go back and configure things and actually at the same before second uh derek houston one of our wonderful super capable students said telecom companies you know they actually did have routing tables for running their calls and routing and phone things basically had a dial plan which was a lot like a routing table that determined about calls from one call to another and you know what then there were these competitive local exchange characters that took over the big players and then it all went voice over rp anyways so or at least a lot of that so that's kind of the thing and definitely derek you're there amram back to you emory you're speaking to a mute i think could be wrong one second yeah so just to add here uh as mike uh this is like a data center this is will be a like as 500 would be your cloud provider aws can be another company uh the reason we show it to you so it can be any use case it might be a gcp it might be azure uh so basically you are you're interconnecting with different areas things different companies different organization maybe within your organization uh there's a separate network so and maybe on the left hand side you're having multiple connections with data centers the reason we show you different use cases just try to simulate the network piece as long as the cloud concept is how the routing works how you manipulate ipgp vgb uh who is doing what what are what are the rr uh router factors i think that's the last step i have my configuration perspective where we'll make r3 r3 as a rr the rr is it's just like you can figure out just like in your transit and when you don't want to have a mesh connectivity uh the rr could be used it's like a hub and a small kind of a scenario okay so let's play with that route reflector concept a little bit more so ibgp is non-transitive meaning if i take a route from imran i won't give it to my wife lisa and therefore imram can't communicate with lisa through me great right but what if i want imran to communicate through me i either have to fully match my ibgp peers which puts us into this scalability problem the number of ib gp pairing session you're going to need is going to follow this formula n times n minus 1 divided by 2. so 100 ib peers 100 times 100 minus 1 is 99 so 99 times 100 divided by 2. can you see exactly how that will not scale very fast it will become a mathematical nightmare so with aws you have cloudhub basically taking your outpass into somebody else transit browning where did cloudhome come from what have we done 30 years ago we had a route reflection what is a route reflector it says do not follow the rules of non-transitive routing now i learn a route from imram on my autonomous system i pass it to my wife liso now imran and lisa can talk to each other so that's what a route reflector is perfect so i think our our last uh step for the configuration of course i can wrap up that perfect and then we're going to start playing with it and tuning it and really exploring the traffic flows my fun time so uh on r3 i'll configure this router so on an r3 i'm telling uh this is a basic conflict you need it uh uh i'm telling the neighbors 102.2 router 2 and router 4. you are my or the client basically now now you're seeing the things are coming up the neighbor was not open earlier so because we definitely see the neighbor so what happens they form an adjacency you get a syslog message and then you start noticing things the neighbors are up and now we're gonna have some cool routing sorry you have to excuse my enthusiasm i've been playing with robbers and switches for 25 plus years almost 30 years and they're just fun for me i'm like a little kid with my propeller hat once we get a network router or switch up now he's making sure the next hops are reachable by making them the self and once we get all this routing configured we're going to check make sure it's up then what we're going to do is we're going to trace the path of the traffic through the network then once we trace the paths we're going to show you how we can tune those paths step one at a time you guys having fun building a network in real time if you're having fun type network fun in the chat box i'm having fun something is broken here if we're missing a route or a route back yeah so let's figure out where we lost it leo network fun nick love network fun original network fun and good say oq network fun jam you got it peter hunt you're just starting and that's okay we'll get you there network fun from derek houston awesome amaranth i know i wanted to get you learning some networking amaranth and awesome chris johnson welcome alonzo bad storms if you lose your power i understand i live just outside of palm beach florida and let me tell you i have a network storm of the day i have four ups's running in my house one for my network here one for my server room one for my mac and all my monitors and i've got another one running too so network fun guys love it you know i made some intro to bgp videos would you guys like for me to do like a one hour intro to bgp video coming up in the future and then for example have a network expert demonstrate all these things is that a desire if so let me know intro to bgp in that chat window now i'm going to go back to imran now these because he's been doing some debugging and things you're good at are you if you want to yeah i'm i'm good i think we need to restart something here i need to step away as well or something or return to mike can we wrap it up here or okay i want to continue um ideally i'd like to actually show this through the traffic engineering but if you have something different going on i understand completely [Music] well we will do as much as we possibly can imran and things change what we'll do is we'll have another session coming up soon absolutely we haven't just finished the basic setup and basic routing setup and listening to need to troubleshoot which need to be maybe something is very small missing and then we can continue the further session i think in bgp there's a lot to offer uh as as i said uh the bgp we just configure ebgp multi-hop we configure the authenticate we haven't configured the authenticator is very simple we configure ibgp router vector uh but we still need to have a route filtering and as you just described yes path route summarization we just play one attribute with a origin we have to play mad weight and ears path as well and advertise export community i i already showed them the multi-path thing and of course the redistribution uh router vector we still need to have uh some bfd uh thing route configuration and of course uh some end-to-end testing further and showing the audience the trays and everything absolutely so we've got the basic routing do we have enough to do trace routes through the system or do you need to go back and and we want to do that another day i think we want to do we'll i'll just save the config so mike and we'll continue with from where we are right now i will not touch it and we can troubleshoot all together i save the conflict of the devices and and can start from where we are right now okay so that sounds okay to you totally fine so this is what happens in life you can be there with your lead architects you got a plan you're planning to do things and certain things pop up on fire so here's what we're going to do imran you want to do this finish this next week and do a second series absolutely we'll start we'll start from here and i'll save all the config you'll share the config with the audience as well later on and i will save this lab and do the start where we are we stuck here right now okay perfect so we'll do the following i've told you guys that i would stay as long as possible up till five today so with em room what did we do we started building some networking things then we configured some ip addresses then we started configuring some routing protocols don't go anywhere anyone i've got you covered after we did this we did a little bit of looking at routing tables so it seemed to me there was some a little bit of confusion about certain things in the network so let's do this let's have amram and i we're going to come back next week and we are going to have a really fun broadcast section and in between now and next week i will stick around for the next hour because i told all of you that i would be here from one to five and we will be here and if you want i will do some live network training we can talk about routers we can talk about switches we can talk about interfaces on devices we can literally talk about any network concept you want amram like me is highly requested and highly in demand so if he needs to go and do something i'll go back and play network teacher for the rest of the day and i am more than happy to so if you want to do some intro to networking type intro to networking and i'll start doing it now we'll come back with interim next week and we will do an advanced series of this we're going to take this base thing that we want to do right now and from here we're going to go deeper and deeper and deeper but if you guys want networking i've been in networking forever and i'll i'll walk you through it yeah i think that the key thing is uh aside from troubleshooting i think that would be the part of our ccia okay troubleshoot let's troubleshoot what went wrong so we'll troubleshoot with the people we'll simulate this again the key key thing is we still need to do is path manipulation how the communities work and how do you park manipulate uh [Music] differently and that's that's the different different communities and different uh path for bgp available that that is a very interesting part maybe you use the same topology where we use another topology but we'll continue to that in apology uh custom escalating um we have a live customer as well uh all the time and customer escalating for something and i i booked the slot for three four hours but some something someone presented so quality awesome great like i said in life imran things happen that's why many of us come from a world where we're taught to adapt improvise and overcome so i'm going to do some intro to networking amram every time i speak to you which is pretty much every day i am always grateful and thrilled for the opportunity you brought so much today and i am so grateful for sharing your knowledge of the community i'm excited to know you i'm excited to work with you i gotta tell you you may see some really scary deep networking things that they're at the cci level talking about how to become a cloud network architect or a network architect and you know what i'm going to see some things coming from mrmeni too but now because imram has some things that he must do um things happen last minute i will do it guess what it's a really big day for me three days ago was my birthday today is my absolute 25th wedding anniversary i am married for 25 years my wife is still doing gymnastics because she does gymnastics like a 13 year old and i think it's kind of cool and you know what now is time for networking so emra might want to thank you for all your help i was truly grateful for your experience i will give you a call tonight before i take my wife to dinner for our anniversary and in the meantime thank you so much great thank you mike and reach out to me if any question i can respond right now i sure will i'll call you tonight um thank you again for your help you can um do what you need to because when fires have to be put out guys like you get called by the world's largest organizations when they have network outages and say fix it right now i know you're a firefighter of the network so thank you do everything you need lots and lots of appreciation lots of respect let's do some intro to networking stuff everybody we'll do it on the fly it'll be a little informal but we're going to have fun okay thanks bye so okay let me see how to go back to this i'm not mr zoom so um i'm gonna unshare everything chris can you make sure that i can share my screen again you should be able to now okay so let me do this um how do i go to the center of the screen um do i need to do anything because all i can see is i don't want to i'm just going to pin myself now if i pin myself does that work chris you're in the middle of the screen i'm in the middle of the screen i don't need to worry about it okay crazy chris is my chief operating officer chris is the super smart person that makes sure that everything we do on the back end works so it's a huge thank you to chris now let's talk about some networking things so realistically speaking when we're talking about networking we're talking about communication in fact we're actually talking about some communication which uh exists between point a and point b so let's work on it let's talk about it you know people wanted an intro to bgp so do you want a bgp lesson you want me to do this live bgp lesson where you guys can ask some questions about attributes and things we'll do it right now on the fly we'll see how fast i can do a bgp lesson you guys want intro to bgp right now you guys wanted at some point i was going to do it on a different day amram has something going on today he puts out fires you want to learn about bgp right now because we can have some fun we will talk about bgp from the beginning the middle and the end you guys all ready for bgp let me know by typing bgp in the chat window and we're going to go straight into a good one-hour discussion on what is bgp how does bgp work and then when you are ready next week when everyone comes back we are going to do all this advanced manipulation that i'm talking about you guys ready type aws bgp and let's get started well the singing when you meet chris you're going to realize what a great guy he is he does so many great things for my students he's a data scientist with an mba he's really really sharp smart so let's begin here i am mike everybody can see me for five seconds okay thank you for coming on chris this is my chief operating officer chris johnson he is behind the scenes he makes everything work so great thank you so much chris so what is bgp so we're going to talk about bgp and bgp is a routing protocol most specifically it's an exterior gateway protocol and i'm going to get a little nerdy here i'm going to get a little in the weeds here and i'm gonna have fun with it so i used to have this hat with propellers on it because when i was in school somebody called me a geek no i was an expert martial artist and somebody called me a geek as an insult they called me a propeller head so you know what i did the next day i went into class with a hat that had propellers on it so i've been a little bit of a propeller head ever since so that's how we martial artists deal with threats we basically say this is who we're going to be so bgp is an exterior gateway routing protocol and we're going to get really technical here my apologies what is bgp it's a routing protocol and when you want to learn routes you've got multiple choices for how you do it option one is as follows you can statically configure it you saw how we had mram do of static route ip route 192 168.0.0.255.255 that's zero to zero 1.2.3.4 is the next hub well you could do that or as you saw we configured ospf internally and ospf we use internally because it gives us fast and absolutely rapid access to our information which is great but what happens if we need to scale what happens when we can connect to other people with lspf we've got no route filtering for the most part we've got speed we've taken all these routes from all over the world we've got to run this spf calculation following the digital algorithm so when we're doing this we're taking in a lot of information it's too computationally expensive with an igp so where you're using an exterior gateway protocol such as bgp for this so that's why we use bgp now bgp or routing protocols build a map if i wanted to go to chris's house and i know he lives somewhere near the tampa region i think at st petersburg whatever i live in the palm beach region i need a map to get there so i could do the following thing here is what a static route is i make a map that says take i-95 up to a certain level and then make 11. and then go west then go right then go left but i've manually defined all of these steps hard one by one this is a static route go here go here go here go here go here so um that's what a static route is by comparison you've got a gps in your cart and the gps says make a left here make a right here recalculating recalculating recalculating nope go straight that's why we use a dynamic routing protocol such as bgp we want something to calculate stuff on the fly we want to redirect our traffic on the fly that is the point of bgp so bgp is used for this now most of these routing protocols and we were talking about ospf and they say hello hello hello are you there they use a hello and those hellos are typically ip multicast and yeah and they're multicast because they want to identify everybody on their subnet why do you use that with an igp you got to figure out who everybody is and what they can do now when we're talking about this connecting to external organizations now we're in a different layer of fun we've got a filter if i get information from somebody i might not need all their intervention if i'm connecting to the internet did the internet need to know the innermost workings of the most private most critical subnets of our network the fact where our intellectual property exists of course not if they did they could reach us if they can reach us they can hack it so when we're dealing with connecting and robbing we've got to use the right routing protocol for the right reasons at the right time kind of like prescribing the right medication to somebody the right dose the right patient right medication well when we're dealing with networking got to get the information so let's talk about bgp bgp is a path vector protocol remember when imran was there and you were seeing the autonomous systems numbers that's the path now bgp is unlike a multicast running protocol uses tcp which means it's reliable one router forms a neighbor adjacency they kind of establish a relationship and then they start exchanging information now they have different states when they load information and we're going to talk about that finite state machine soon but at least know what it is and yes i'm a little enthusiastic when i talk networking because i love it if you don't have a network nothing works so want a good cloud computing world learn the network so bgp as i mentioned is used with scalability and tunability and traffic engineering so let's look at it so when we're dealing with bgp we have a few messages there's an open message there's a people like message there's an update notification update message and notification message what are these messages well let's talk about them and how this works so as soon as i established a tcp relationship with the other router in my bgp hearing session we become neighbors so what happens we as well first we do establish a tcp session and how does this work with the tcp session you know you send us in at the far end the frn basically acknowledges your sin sends a new sin to you tcp message is established thank you genie very much um 25th anniversary really excited so so we establish smash the tcp connection and then after we establish this tcp connection we send an open message and the open messages that routers are sending each other are identifying themselves hey guess what we're using bgp version 4. my autonomous system is tuned is 200 and guess what these are the timers we use so what do i mean by time risk you can tune these timers remember how i like to talk about load balancers and i like to say a low balancer says with its keep alive are you there are you there are you there with a self-check message well guess what with tcp you're sending messages and make sure they're there too called to keep alive so what is a hold timer basically what happens is you form a relationship you send an update message i'm an open message in the uptown message you determine what form of bgp you're using right now everybody's using bgp version 4. you send your autonomous system number and you send whatever timers you're using meaning a neighbor goes away for example if you can't find it after a certain number of hello messages remember keep elizabeth the whole time is basically how long it is basically three or four people like messages that are missed so basically establish tcp session send an open message which identifies what you are able to do and your capabilities as a router now after that after the session is done you've got to basically keep your session up so how do you keep your session up and everything else they like for example with the health check with a low balancer are you there are you there are you there you know what you're doing here you're basically sending a keep like you're there right you're there right you're right right and basically speaking there's no message to come across in a period of time the name the peer doesn't send a keep alive to say i'm here i'm here i'm here bgp neighbor is taken away and all the routes that were learned via bgp all go away so they all go away so what we're going to do because this bgp stuff is really complicated i'm going to talk for about 10 minutes then we'll have a couple minutes of questions i'll talk for 10 minutes we'll have a couple minutes of questions so the next message that you need to understand is something called an update message so remember imran configured a bgp session so first thing that came is we sent we had a tcp connection then we sent an open message then we have keeper lives going i'm here i'm here i'm here don't forget me that's how we know now let's say a route that was in our routing table went away well if that route's not here when we continue to send traffic to that destination we have our traffic dies we want to withdraw that route so we can take another path remember that recalculating recalculating we're calculating your gps we have to have that happen here too so that's what we're going to do so we're going to recalculate we calculate and how do we do that if a message is lost for example or a neighbor goes down we're going to send an update message which basically pulls old things and guess what if we learn new messages we're going to send an update that says i've got it i've got it i've got it so that's the message so let's go back to these messages open basically says hey by the way i want to form a neighbor relationship with you this is what i've got going on after the neighbors up keep alive i'm alive i'm alive i'm alive guess what i'm still here update message you gain a route lose a route withdrawal route got a route you send your neighbors an update guess what here's the new route i learned now after that there's something called a notification message a notification message is not a good thing a notification message is sent where something bad goes along the way if you get one of these notification messages it's not a good thing so you're going to be doing some debugging and guess what when you're dealing with routers and switches or any tech you're going to be debugging debugging debugging so let's talk about that now i want to talk about how bgp forms a neighbor relationship and we're going to talk about this machine but before we get to the machine are there any questions on bgp for example what it is why we're using it i like to start at the elementary levels and since i have the opportunity to do today we're doing it on the fly so it's the 25th wedding anniversary special of bgp on the fly because amram got pulled into a firefight which is totally cool thrilled that he's doing that helping his customers that's what good architects and good engineers do any questions for me right now i'll pause for a minute to see if there's any questions i'll type in the chat window because there's about a 90 second delay and if they're not i'm going to keep going into bgp but i want to make sure we enter everything i want you guys all to do great typing in the chat window because i know how long the delay is i'm going to wait for about 30 seconds 45 seconds see if any of you wonderful folks out there have any questions on vgp and if not i'll start talking about the finite state machine chaitin nayak let me answer your question here these are great questions keep them coming do you have to manually configure the routers to translate these messages and take action especially for notification messages the second you configure um a bgp session automatically these messages occur so basically these routing protocols are just algorithms and they have a set thing to do so as soon as we establish a neighbor relationship the open message is immediately sent after the the neighbor relationship is established to keep lives happen automatically if we learn or lose routes those updates happen automatically and if we get a withdrawn message or we get that kind of notification message it's because something bad happened maybe the tcp connection was shut down for example or um we ran into a problem somewhere else so chintan i hope i answered your question um um generally speaking notifications are not a good thing mr p after bgp what's the next big thing well realistically speaking i don't really think for a while there's going to be nothing other than bgp bgp is the only protocol that's really used to track to connect to external entities i'll tell you what other things are coming though mr p um and mr p i gotta tell you i've seen the speed that you learned and i've seen your ability to multitask and it is impressive i meant to actually write you a note after seeing how impressive you were in monday's class and the way you were actually and the environment for which you were actually learning it was unbelievable so bgp is realistically the routing protocol that is used to connect to external entities now is there anything else going on with networking sure there's something called software defined networking what software defined networking is and it's going to come into play soon typically speaking on these routers you've got the control plane which are the things that determine how do you move your traffic what is the control plane it's ospf it's mpls speed you've got a data plane or the forwarding plane which are your routers and your links with a cloud guess what you've got your servers and your storage that's your data plan you've got these hypervisors and container orchestration services that's the control plane same thing old network old technology and a new environment out the cloud solid cloud now we're one of these things with the cloud so i do see software defined networking what is the difference between software defined networking and bgp and traditional networking traditionally all of this stuff is on the router or like all of it's on the router so you want to learn on what's on the router you go to the routers now with sdn you've got these controller modules on these big giant servers somewhere else and these big giant controller modules are building maps to the network and they're looking at things such as links bay link congestion latency and sdn your traffic will be rerouted around the internet into an environment that gives you near direct connection like performance remember internet performance is typically not good why is internet performance not good i'm going to tell you right now you have guaranteed access to your speed to your internet service provider but once your traffic gets on their network you've got no guarantees if your network has to go through 100 service providers to get to its destination guess what your network goes through two internet service providers and after it leaves your eyes there is no guarantee of anything that's going to happen with your traffic so because of that you know you've got nothing and that's why organizations use private lines or direct connection instead of vpn but with software-defined networking these controllers which are off of the network can literally take your data and reroute it based on latency speed performance lack of conjecture and give you private line or mpls like performance over a public network so you will be seeing a lot more sdn coming sdn is bad at its very infancy stages bgp is going to be used by 99 of organizations to connect to the cloud for long periods of time in at least a decade and maybe more but this sdn technology is getting better my friend kumar mehta for example the one on his brother of pervimeter the founders of versa network have a very good sdn solution cisco has got a very good sdn solution juniper networks has got a very good sdn solution and there are some new and innovative providers that are doing this so that's where i see the future bgp for a while last mpls and label switching like people like me that have done traffic engineering but now we're starting to look at sdn and we're taking it real seriously so mr p i hope i answered your question let's get back to bgp and you guys ask more questions we'll do it but i've been told by my person my team that i should actually act like a youtube person so if you're enjoying this video if you can please leave a like or subscribe and if it's appropriate and fun if you want to forward somebody else to this um we would absolutely love um to kind of work with you guys so let's go back to the neighbor relationship so when bgp forms a naval relationship it goes through what's called a finite state machine um what do i mean by the finite state machine there are things that actually occur through the entire steps along the way so as follows we get an idle message we got a state called idle connect active open sent open confirm and establish so we'll walk through these and then we'll talk about how to tune traffic that's where the fun is the first stage is idle so basically speaking i've got two bcp speakers i put neighbor 1.2.3.4 remote as400 and instantly i come up as idol now what's going on is idle my router's up the other routers on the far end we're trying to form that neighbor relationship until the tcp connection is established we are idle no if we run into any errors along the way misconfiguration mismatch of bgp versions or something like that we will go into idle and stay in idols we will never leave idle idle is a place where we want to be when we turn it on but if we stay there for more than a second or two this is bad news something's wrong so what is the next state as soon as we start to form a tcp connection we enter connect and what happens the bgp starts the connection the process basically opens up and sends an open message to the neighbor and guess what we transition into open center this is what's supposed to happen now if we can't establish a tcp connection we can't send an open message to our neighbor because we don't have the session we can transition to open set so if the tcp connection does not go up we've got a problem what's going to happen is bgp is going to try and it's going to go to a connective state active is not good it should go to connect open open send but if it goes to active something happened along the way so here's what's going to happen if the connection doesn't go through it's going to retry in the connect state again and again and again and if we can't find it guess what's going to happen it's going to transition back to idle so you should start at aisle and then you should go to connect and then open and open sent this is what happens when things are right when things go wrong and when we're talking about tech things always go wrong they always go wrong which is why you know just knowing a name of something is never enough you need to know how it all works especially when it comes to troubleshooting so that's what's going to happen it's going to go to connect open open sent all things go well but things aren't working the way they're supposed to be don't you worry it's going to transition back to idle and you're going to have to go figure it out so what's the next day so we already told you what it should be but remember i said if you don't get the perfect tcp section or things don't work right we're going to transition to this active state okay let's talk about active and active transitions in the active state what realistically happens you keep trying to reset the tcp connection for a few reasons you keep trying to start the process if you can't you go back to idle if at some point in the active state the tcp connection comes up and you start exchanging messages you gotta open an open set the way it's absolutely supposed to um yes muhammad we'll talk more about sdn in a little bit because there's many different definitions but it all involves the separation of the control plan and the data plan for forwarding so now let's talk about what should happen idle connect open message open sent because it went through now we're in open set so what are we waiting back here for now we're waiting to get an open message from our neighbors if we get an open message from our neighbors it's like okay this is really fantastic we've initiated connectivity to them they've initiated it connected to me ding ding ding everything's working really great so now if anything goes wrong we go back to idle so idle's normal in the beginning it's what it's supposed to be if you're an idol after a minute later there is a problem go troubleshoot go debug so when you send that open sent and you're waiting for the open message back when you get the message from the far end you're now in something called open confirm basically idle connect open open sent confirmed you know what confirmed me means you received information you're waiting for the keep a lie from the neighbor you get that keep a lie from the neighbor boom it's like magic you now transition to established and when you're now established and that bgp neighbor is established like memorandum was talking about you've got a full neighbor connectivity this is where the routing information goes this is really great this is everything the way we want it to be open sent open confirm established life is good in established we are up we are running and this is where things go so now we're going to talk about attributes these attributes that we're going to talk about next are where all the magic happens in bgp so before we go to the magic of bgp we discussed a little bit of complexity about tcp opening up the session so let's recoup very briefly we talked about bgp being an exterior routing protocol that's used for scalability tunability and traffic engineering we talked about it being unicast using tcp port 179. we talked about why we use it what did we talk about next we talked about the four states of messages between the open message which is basically establishing the tcp connection the keep a live message which is basically saying i'm there i'm there i'm there i'm there i'm there i'm there yes we're still here together we're at the party we're going to exchange routing information an update message which is i've got a new route or i've lost a new route and a notification message and what is that in the notification message the notification message is as follows something went wrong removed it for example so now you know the four messages now let's go back to that finite state machine and what's supposed to happen what's supposed to happen is you send you're in the open state well actually you start with the idle state and and uh realistically speaking you start in an idle state which is what is exactly what happens then after the tcp connection is established you go to connect then you send an open message and if all those wells you go to open send and if all is working you get to open confirm you wait for a keep alive and if everything goes right you are established ding ding ding everything's working the magic's happening anywhere else you're reaching everything what if you don't establish that tcp connection what if you go to idle and you try and go to connect and instead of sending that open and getting to open center something bad happens you go to active active is going to keep retrying the tcp connection and thing if it doesn't get established it'll transition back to idle so idle is not where you want to be except initially if you stay in idle you got to fix something so prior to getting any more because some of this stuff is pretty complicated bgi bgp is a cci level routing protocol and unfortunately it's the routing protocol that you need to use to connect to the cloud why do you use it to connect to the cloud cloud's exterior they're an external organization do you really think aws could take a million routes from a million and one companies and still put them in there and a nine-to-one interrogate with protocol of course not and that's why they limit the routes they need to take you is a hundred routes meant miniscule and incredibly small yes is a hundred routes from a million organizations on aws unbelievable number times the hundred times a million it is so when you start looking at cloud providers you got to look at the architecture of why they are what they are why do you have to limit certain things because they've got limitations because they are so huge so if you hear me say saying small amounts of routes get creative they have to have a smaller amounts around how could they do it any other way so can you get better performance in the network of the data center of course you can because you're only dealing with your stuff but look at the magic of the cloud provider look at the ability to handle thousands and hundreds of thousands of customers on a shared network media and still deliver some pretty good performance and still be able to deliver some pretty excellent security and be able to deploy something in seconds versus minutes while you're on to something is truly transformational so understand the strengths and the weaknesses of all approaches that's what makes you a great architect when you know all your opportunities and all the things you can do to improve your customers you're a great architect so muhammad yes every vendor has its own definition of sdn sdn could be a little bit of something like a cheap router with a commodity router or switch with fancy software that's cloud managed but generally speaking when you're talking about software defined networking in the world that amram and i come from the juniper networks world the cisco networks world the really big verse in networks world the people that are doing 80 percent of the core internet routing what we're talking about is separation of the control plane for the data plane and using the separate control plane to really manage your traffic based on intelligence that it finds in the system so i'm gonna hope i got to your question and any more questions i will also type in the chat window because i know there's the like part of moving on to the next section so let me ask this why am i typing in here because that 60 to 90 second delay i don't want any of you to miss any opportunities whatsoever anywhere any time to learn so please understand that so while we're at it i'm going to check and see one quick thing any questions for me if you've got any questions we want to answer them we want to be a source of information to help you in any way we can your cloud computing careers and network careers are really important to us mr p since you have to manually connect every peer connections on average how many are we talking about wow what are great questions so it depends on the complexity of your network misrepair for people like me that have worked on these large isp networks we could be dealing with thousands upon thousands of routers that all are doing ibgp pairings together with on our own networks with multiple route reflectors and something called confederation is literally thousands of them when we're dealing with the cloud here's how simple it's going to be we're going to peer to aws maybe for example we may peer to uh azure we may periodic gcp realistically speaking that's three sets of peering session now anytime we're using direct connections we're using peering on vpn connections we could be using pairing anytime we're using cloudhub for example or transit gateway mr p with aws or the azure equivalent or the gcp equivalent you're going to have bgp pairing sessions so what is cloudville and we'll map it out soon cloudhub is a way that you can use ebgp connections to create a hubspot routing environment so if you had a hundred vpcs and you're using cloud hub you'll have a 100 vpc pairing session if for example you were going to fully mesh your ibg prepares because you weren't going to use caught up you and your vpc pairs you would have a whole lot of bgp pairing going on in your environment so what can you con what can next question for me mr p what can we use to map and confirm these connections well on your router you'll be able to deal with show ipvgp neighbor and that's all you're really going to need to see on you'll be able to connect it from the routers to your things now when you're gonna because the reason i say check from the routers when you're gonna connect to aws they're gonna have their own thing now they have their own virtual routers and therefore you can easily see your routing tables the cli or the management console but when you're really talking about getting into your bgp tables and really see what's going on when you want to really look at these messages go to your router the cisco router will give you a million times more diagnostic information on your connection than you're going to get from the cloud provider why it's all the router does the router only has to do one thing and because it only has to do one thing it's going to do it better and faster than anything else the cloud provider is amazing they've got a control plane layer of software that's literally doing anything network virtualization server virtualization container virtualization storage virtualization i mean they're doing a lot so when you're a jack of all trades you can never be as good as something that focus on it exclusively so when i need a fire while i get some from cisco or checkpoint or palo alto or fortinet i use an industrial grid firewall put it on an ec2 instance because that's all they do but i do this for super high security environments in modern environments i use the cloud native services when i need real heavy network stuff it's going to be on the network itself so mr p show ipbgp neighbors and that's going to show you all your bgp peering sessions hope i answered your question there mr p who else has any more questions does anybody have any more questions before i go on to more content if not i will go on but i'm going to wait 30 seconds to see if there are any more so while we've had an audience there's been over 4 000 people so far that have participated today which means we are excited and thrilled to share our knowledge and create the next generation of future cloud architects and network architects i'm going to get back to the presentation i've done my 30 years in tech almost 30 years it's time for people like imran and me to create the next generation of leaders we're not going anywhere we're still here we love it we're not getting out of this but you know somewhere along the line we need to train newer people so that newer people can get involved and have the magical careers we did so that's what we're doing these sessions so if you need any kind of session you want any kind of session you want me to produce content tell us in the chat box send us an email to ask cloud mike i will run it through my production team and we'll research the concept leo inside of an organization what requirements will define whether if you use ibgp or ospf that is an extraordinarily good question leo inside an organization you're always going to use an igb and igp is used for rapid speed rapid convergence so you will always use something like ospf inside of your organization now when you want to connect your organization to an external organization leo that's when you use bg create so ospf is used internally bgt is used externally so realistically speaking think everything inside of your house uses an interior gateway protocol because it's inside of your house if you want to connect your house to somebody else's house like you want to connect let's say your house and your network to my house and my network via vpn and then you want to run bgp to route between with bgp i could give you access to say my vpn subnets and my demilitarized zone and i could not give you the information to reach my production subnets where my information is stored that's why we're using bgp to connect and filter and scout when we connect to external organizations so internally ospf or eigrp or intermediate systems to intermediate systems externally like between your data center and the cloud bgp externally between your data center and your internet service providers ebgp hope i answered that question why would you use ibgp there's only one reason to use ibgp if for example i learned information from a tnt and i also learned information from verizon and i want my network to know my whole network internally didn't know which paths verizon does better than an att versus which ones atm t does better than verizon i would run ibtp across my networks my internal network would know i probably wouldn't do this in real life i'd probably just introduce the default route into my igp but i uh igps like ospf inside your network egps just like for external networking and that's what you need to know and that's how it all works so hope that makes sense so now that we've gone through the finite state machine and we've gone through how we set up a connection let's do this let's now talk about the bgp attributes and then if you haven't seen this when imram and i come back next week it's gonna be super super fun and will actually make sense so their bgp has something called an attribute and an attribute are simple knobs we used to call them nerd knobs an attribute is something about the route that we actually learn so maybe we learn the weight maybe we learn the number of paths it's referred to so kind of this is what we talk about when we talk about bgp attributes so i know bgp is fun for me i know i get excited when i talk about bgp but i gotta tell you i love this routing protocol it can do so many things so now we're gonna talk about the attributes so there's a bunch of different attributes remember today we worked on the origin attribute okay what is the origin the origin our attribute kind of tells us exactly how we learned the route did we learn it from our igp this is preferred did we learn it via bgp this is not preferred or did we just redistribute it connected and not set a metric in which case it's going to show up as incomplete so attributes remember learn from your igp most reliable learn from your etp second preference incomplete not good last choice so when imram made that route map today and in that route map what he actually did is he changed the origin code from incomplete to something else well then you know these things happen so let's look at the next one it's the next top attribute what is the next hop attribute let's talk about it so the next top is the ip address where you would send your traffic to next to reach the destination i'm going to tell you everything that doesn't work in vgp most of the time it's because you can't reach the next hop so when amram was doing a show ipbgp he could see the bgp route and the next hop when i was doing a show ip route i was seeing which routes that were learned from bgp replacing the router's routing table which is going to determine where our traffic is forwarded there were two different things he's looking at the bgp table if it's not in the bgp table it can't be in the routing table and then the routing table shows up so the next top is where we're going so let's talk about this first attribute the first attribute was the origin code hey how did you learn it the next attribute is the next hop now these are where it starts to get fun and the reason this starts to get fun is this is where we can do our traffic engineering so these next things are really really important so the first attribute is called the weight what is the weight weight was a cisco proprietary attribute but aws and others support it as well known weight is as following the higher the weight the more preferred the route is that's it it's an attribute local to the router the higher the weight the more preferred the router route is we'll talk about more about that in a minute we'll walk it through next and so we're going to prefer the highest weight if the weights are now equal guess what else we're going to prefer the local preference so prefer the past with the highest weight this is the decision algorithm next prefer the path with the highest local preference now if the local preferences are the same remember this this is super important in fact i'm going to actually paste this because it's so important in the chat window so we can do some okay that's not going to work as well as i thought it would so let's try and do this so this is what happens when you are not mr powerpoint and so let's do this so i'm going to let you know these are the attributes in order prefer the path with the largest wave next if the weights are equal prefer the path with the largest local preference hmm if the local preferences are the same take the one that was originated locally on the router now if all of that is equal we're going to take the one with the shortest as path shortest number of hops so that's next that's why i put it in the chat box now if all of this is true then we're going to use the one with the lowest origin code remember igp is preferred over an egp and that's better than incomplete and that's why imran when he redistributed connected subnet added an origin code so then where are we coming from if the origin codes are the same then prefer the path with the largest meg or the lowest med multi-exit discriminator the lowest metric to leave our autonomous system and if the meds are the same prefer evgp routes over ib gp rows oh wait wasn't the administrative distance automatically lower too when we looked through that before now if the routes are equal we're going to start getting in the goofy stop for further pass with the shortest path to the next stop okay so they're there like i said like muhammad weight is only locally significant and it's not available on all routers it is available on cisco routers it is available in the aws environment and certain other routers as well but it was cisco proprietary for ever then now if the routes are the same and the reason i'm giving you guys this is we're going to work through some scenarios and i want you to tell me which one would go so let's do this so then you start picking stupid stuff in the end of the equation like lowest ip address highest ip address so now you know kind of where we're at so you know what bgp is you now know the decision algorithm and you know some of the attributes so what are we going to do with this well let's have some fun with it now so let's uh do this let me let me work through some slides with you now so we can take some situations okay bgp bgp can you guys all see my desktop with some powerpoint slides right now now let's go and i want you guys to all look because we're going to be tuning traffic engineering now in this particular environment we have our data center connected to our cloud now we're going to have fun with all that cool bgp stuff we were talking about bgp stuff this is fun so because of this what we're going to do is as follows oops we're going to do this we're oops i don't exactly know what happened i'm there that's why i'm not mr powerpoint but that's okay so we've got two connections let's say we have two direct connections to the cloud now i'm going to tell you first before we even get to any kind of bgp attributes routers always prefer the most specific path to a destination meaning if you've got a slash 15 and a slash 16 you will take the slash 16 over the slash 15 because it's a longer prefix list which means it has more information so now you get it so in this situation no i've got two direct connections to the cloud i have not manipulated bgp policy at all on the top link i am advertising a specific subnet to 172.16.0.0.16. i am advertising a summary room which covers all the cider range of the data center on the top link on the bottom link i've got i'm advertising a 172.17.0.0.16 and i'm also advertising a 172.16 up 172.16.0.0.15. okay what have i done i've got a specific route on the top link i've got a specific route on the bottom link and i've got this summary route that covers our wholesaler range under normal circumstances which link is being going to be used for 172.16.0.0.16 tell me in the chat box am i going to use the top link or the bottom link knowing that routers prefer the more specific route let me know in the chat box stop link or bottom link who are we taking to 172.16.0.0.16 architecture in action someone out there mr p muhammad somebody's got a networking background or even one one of the cloud computing people which route has a more specific route which link is going to be used for 172 16.0.0 let me know in the chat box peter hunt excellent we're going to use the top link anybody else we're going to use that top link for that 172 16.0.0 because it's more specific does everybody see that guys are awfully quiet it's not like a zoom meeting where i can see your faces or talk to you so please let me know and if we're taking 172 16.0.0 for the top link john thompson great job anybody else great job leo great job okay i know it's a little tricky everybody nick love great job top lane excellent you got it 172 16.00 16 is the top link not mr p excellent good job everyone you're making me happy it's gonna be some heck of a 25th anniversary three days ago was my birthday i met my wife at 21 married her on my 23rd three days after my 23rd birthday and today we've been married for 25 years so anyway i'm just excited it's a good day and you guys are getting bgp which is the most complicated and ugly of all the routing protocol to learn so i'm excited now 172.17.0.0.16. can somebody tell me which length that's going to be the primary link under normal circumstances please 172.17 which link is going to be used which has the more specific link for 17217 i know there's a big delay so i'm going to give you a little bit of time tell me who which link is more specific for 172 17 so 0 to 0 16 the top link or the bottom line leo pratt is great bottom link absolutely yes yes yes great job bottom link is more specific to the 17217 john thompson excellent nick love excellent you guys got it now does anybody know why i'm even advertising this cider range of 172 16.0.0.15 gold star for anybody that knows this one if not i'll tell you mr pig nick love alonso pig hunt reginald great job you guys are getting it why am i advertising the 172.16.0.0.15 amaranth good job r r and rfr excellent job now why why why why does mike feel the need to advertise the 172 16.0.0 15 the entire cider range on both links why would i do that you guys don't know that's okay i'll tell you but i want you to know because it's really really important let me know nick love got it wonderful in case one goes down amaranth's great so if i lose one of these links along the way if i lose a link guess what i have the whole site arrange advertised on the other link if for example i did not have that other link i would not have network reachability so john thompson let me explain this because it's a good question if we have only 172 16 on the top link and we only have 172 17 on the bottom line if we lose the top link and i've only advertised 172.17 on the bottom link what happens is as follows i don't have access to it so what i'm doing is i'm sending a cider range which includes all of it so what i'm doing goes in the top link i'm leaking which is different connection i'm linking on the specific route to make that more prefer on the bottom link i'm linking a more specific route which will be more preferred now that cider range will never be used because we're going to always take the more specific routes but if the top link were to go away the bottom link has access to 172.16.0.0.15. what does that summary route have and mean it means i can connect to the 17216.0.0 16 subnet and the 172.17.0.0.1 so what i'm doing is i'm leaking a specific route to traffic engineer my traffic and i'm sending an aggregate route if one of those links goes away i still want to be able to reach everybody so that's what's going on now let's do it another way let's change the weight so weight is local to the router and because weight is local to the router if we can change the weight of a route we can make it more preferable or less performable the higher the weight remember that algorithm i pasted in here the larger the weight the more preferred the route is so let's take that route with the higher weight and let's go do something about it so what are we going to do on the top link we let's say we want to make the 172 16 preferred guess what we give it a higher weight than the other one and let's say in this situation we take the bottom subnet we take the 172 17 and we don't want to use the top line basically speaking we reduce the weight now we're going to go to the bottom link we're going to prioritize that 17217.0.0516 by increasing the weight and we're also going to reduce the weight for the length of the subnet we want to use the top length the 17216.0.0016. what have we done here we have verified that under good circumstances where everything is working properly we are going to be using the top link because it's most specific meaning it has the highest weight for this 30 things the routing table that's going to be sitting here on the router is going to say guess what i have the best route to 172.17 over the bottom one and the best route to 172.16 on the top link why because it's got a higher weight now why do i still advertise both on both links what happens if the link the bottom link goes away now the top link had no knows how to reach both subnet why because they're both there we advertise both routes and what would be naturally the more preferred one on the lower on the lower link will be more preferred for 172.17 but if that lower link is gone it's dead there's nothing there we still get the two routes from the routing table up top that's why we're prioritizing one on one link and one on another link but we're still using the lower weight for the backup is that clear to any everyone if it is not clear i will go over it again because i want you guys to understand this bgp is cool you guys get that give you guys a minute to ask if you see me going from one monitor to the next monitor like looking in different directions i'm trying to watch the stream on one monitor and then be able to see the questions on another monitor so bear with me so um it's not really an ebgp multi-hop situation here muhammad i'll talk about what that actually is in a minute um what's going here is everybody understand why we're you we're why we have both weights on both links one prioritized one not prioritized in order to make okay so derek yes so what we're trying to do is we're making sure if one link goes away we still have a secondary link and because of that these are kind of one of those things we're looking to do so these are those things that make us happy and excited um realistically speaking so does that make sense for everyone okay perfect it's clear one second okay so now everybody's got this now let's go back to muhammad's question on ebgp multihope okay while the ebgp multi-hop is as follows if i've got one link let's pretend my arm is the link on one side of the link let's say we have a 192 168 1.1 and on the opposite side of my arm we've got a 192 168 1.2 on this typically speaking we would do if this is one sitting 192 168 1.1 we would typically set up a neighbor relationship with the remote as with the 192.68.1.2 that's normal that's standard so what's actually going on here is and that's typically we would have a direct connection so you wouldn't need anything called multi-hop now if you are going to go loopback address to loopback address meaning you have to traverse the wan link add another subnet that's called ebgp multi-hub because it's not direct you basically have to go over more than one connection so look back to loopback connections or why you would use ebgp multi-optim their muhammad so that's why we're actually doing this so now that we've adjusted the weight we're going to have some other things that we can do we can emit the local preference area's path and guess what we're going to but understand how we manipulated traffic first by leaking out more specific routes secondly we actually did so um by what's the word i'm looking for by manipulating the way so what's next local preference let's do that but let's answer mr p's question the blue routers are actual real routers that are actually sitting in the organization's data center office that is connected to the cloud absolutely um the cloud you've got these virtual routers which are logical routers sitting on servers somewhere along the line but these are real routers with real performance so perfect i hope i answered your question so next on the list we've tuned the weight um local preference is next okay we've changed the weight now we're going to change the local preference so with the local preference let's do this so let's let's say on the top link remember the larger local preference is the winner so instead of changing the weight which we previously did we could have just changed the local preference by uh looking changing the local preference lots of things happen so you change the local preference on the top look what we did we prioritized the subnet 1 172 16.0.0.16 with a local preference of 200. we took the lower subnet 192 i'm sorry 172 17.0.06 on the bottom link we flipped it so somebody tell me which route which path is going to be chosen for 172.17.0.0.16. so which has the larger local preference for the route 172.17.0.0.16 which path will the routers take from the cloud to go back and reach the data center for 172.17.0.0.16. which one guys or girls or everyone out there which router which link which has the highest local preference genie which has the higher local preference for 172.17.0.0 the top link has a local preference of 100 for 172.17.0.0.16 the bottom link has a local preference of 172.16.0.0 which one for which has a large or local preference leo paradise you've got it 172 17.0.0.16 the local preference is 200 on the bottom we're on the top link it's only 100. prefer the path with the largest local preference now at some point you're getting with highest and lowest ip addresses and we're going to manipulate the things before we go there pete hunt great leo paredes hunt genie i'm sure you're going to get it real soon he's got the last couple nick glove bottom excellent genie you got it derek great job okay now when amram and i were doing this show ipbgp and we were having a party with it he's configuring things he's doing all that tech piece i'm kind of explaining the tech most of my career has been as an architect you know what half the job of an architect is translating between genius engineers and executives gotta tell you it's half the job so i like translating i learned how to do this when i watch a movie my wife tries to translate it to me there was a period of time where i didn't speak greek when i married my wife and so she used to translate great things to me you know what she does now when she doesn't remember she translates english to english and it's really cute and really fun but all what we do is architects half of it is translating between the genius engineers and the business people so now we're going to prepare to as pass prefer the path with the shortest as path prefer the path with the shortest pass path so given that scenario which has a shorter is path for a certain realm look at the top link we are one hop away from 172.16.0.0.16 and because we prepended or we added aspas which imran will walk you through on another one for 172.17.0.0.16 where two as passed away meaning twice as far now let's go to those bottom links okay with the bottom length what do we see we see as follows a 172 17.0.0 16 with a short path and we see a longer path of two autonomous system out for 172.16.0.0 so from the audience people please tell me which link is going to be chosen for 172.16.0.0.16 and also tell me which link is going to be chosen for 172.17.0.0 please let me know guys right now guys girls anyone out there let me know and if you get this you're really getting bgp tuning in traffic engineering let me know in the chat box okay i'll give you some time i know there's a pretty massive delay tell me which link is going to be used for the 172 16 based upon shorter as path and which link is going to be used for the 172 17.0.0516 based upon the shortest diode path please let me guys know that you're here you're awake alert norian and you're getting bgp by putting in the answers to the chat box okay eric um yes so um which subnet though is the 172.16.0.0.16 that's going on the top and the 172.17.0.0 on the bottom is the vice versa please let me know which subnets are where guys so i know you get it well what's going on sometimes when we see people that are starting to talk about local preference when we're talking about as paths is what's going on is there's delay and sometimes people turn in 60 seconds later they could be seeing different portions of the video sometimes so um peter hunt top link 172 16. stellar job there bottom link 172 17. excellent excellent job peter hunt great job anybody else see the same thing derek you got it derek you once asked me if beyond a certain point in our lives can we still learn er derek you're doing cci level routing on a month and a half of knowing you unbelievable derek reginald excellent job good job everybody good job okay one last metric we're going to manipulate the med the lowest multi-exit discriminator so in this case we've gone back to the same environment and i've decided to play with the med instead of the weight the local preference or the a.s path prefer the path with the largest lowest med the lowest multi-exit discriminator like i don't make up these terms i call things simple i call this a cup of water whether it's a cup a bottle a glass i don't have a million one terms i keep things simple i like it that way it's much easier but you know got a lot of tunes because some things are local some things like the ass test we could pretend it's dependence on upstream so let's choose the path with the largest mid going back to this situation which path is going to be chosen for 172.17.0.0.16. which has the lowest med let me know guys guys are doing unbelievable which length do we choose for the 172 17 by modifying the med metric as stated and this power plant slide somebody help me derek slap 172 16 on the topic because of the reduced med excellent which means peter hunter 172 17 is going to go over the bottom line excellent because it's got the lowest mad multi-exit discriminator amrit good you got it okay this is some good exciting stuff what did we do today we went through this thing and bear with me and we picked paths and we've chosen routes based upon the most efficient routing so we started with you know doing some configuration then we talked about what is bgp we talked about bgp we talked about why is bgp used after we talked about ysb gpus we talked about the bgp messages then we talked about the finite state machine you know open open sound open confirm that kind of thing then we talked about bgp attributes then we talked about tuning bgp so we didn't on the fly entered a bgp class what questions did you guys have fun if you think bgp is pretty cool type aws bgp in the chat window let me know you there type aws bgp i'm proud of all you guys that have participated i'm proud of all of you that have put answers in i'm proud of all you guys that have not been afraid to come off mute any information and just try it you guys are doing absolutely fantastic like me that have been working and spending a lifetime trying to breed the next generation of technology people get really excited and really happy when we see guys we see people learning this fast so can me thank you aws bgp leo paradise aws pgp if you guys are having fun or learning bgp type aws bgp alonzo peter derek amaranth reginald jeannie awesome john richards aws bgt you connor so bgp is cool sarah smith alex jonathan brown you know welcome we're so happy to have you here and we'd love to hear from you alexandra johnson excellent alexandra pressox wonderful bgp bgp bgp i got to tell you you want to work on the cloud obino thank you the johnson 86 aws bgp thank you i'm here with you this is routing this is cloud networking this is things for network architects this is things for cloud network architects and this is things for cloud architecture you're not going to find this critical information covered in certification materials but guess what you need it for the job richard rich excellent so that's why we're doing these things we're bridging the gap between the stuff that certification teaches you and the things that are necessary for the job legal pirates thank you cloud learner thank you kyle j thank you we love bgp bgp is so tunable derek houston my knowledge is thank you so much for these extremely kind words i gotta say i have never imagined how awesome it could be to literally go from youtube share the experiences i've had for the last two and a half decades with the community this is the best feeling you can imagine when you guys all learn things and i get emails from people every day all over the world telling me i've got a first cloud job we are so excited um with regards to um with regards to muhammad's i think you're actually watching a slightly different part of the presentation right now but when we were all synced up during the presentation you had some incredibly valuable things and we loved it all but if you have the local preference that's the same the path is the same arrow then the meds going to come into place absolutely you are completely right so simon cordova there is no news in to donate to these classes we are just excited to be part of the community chris from my team will put out we have a cloud architect career development program which is literally something that summarizes my 25 years of experience of creating architects and helping people get their first cloud job we do it better faster and far cheaper than anybody else that offers a single program i've been in this world forever and i want to raise the next generation of technology professionals so to let you guys know we have all new gear coming here karen will be using sony professional cameras and sony professional lenses and a professional grade of microphones but i found that coming online and interacting with the community see my birthday happened the other day and i received messages that brought me from tears messages from india pakistan asia africa i mean all over south america thanking us for the free content i after i was in tears told my wife showed her the messages and she says oh my god so i basically bought all new sony cinematography gear meaning the stuff they make movies out of and it's gonna be here today and tomorrow next week as soon as we get our gear tested up we are going to be doing a lot of free training in fact we may almost have a live stream show of the day for about an hour we're going to talk about all these things that are current all these things are going on anything you need to know for your cloud computing careers we're going to bring that to you so if you want to be one of our students we'd love to work with you in our club architect career development program our students get the most elite training in the world we can easily charge 20 25 000 for this training my students in the class will tell you that we don't our mission is as follows basically train the next generation of technology professionals at a cost that's basically one to two days pay for people in their fields and that's it we are just a solution provider i've done all the work i needed in my life and now it's raising the new generation of tech so if you guys are excited about the concept of a daily show and more technology please let me know with you know more live streams in the session in the meantime i'm going to ask simon simon cordova's answer you've been programming for six years now getting into architecture you're thankful for this you're welcome programming is great architecture is the best job i have ever had i practiced internal medicine first i became a network engineer which i loved i mean i really enjoyed it and then i became a network architect first and let me tell you that meeting with customers that designing of technology solution like solving customer business problem oh my god best experience and when i went in there and i focused on my executive presence and trained soft skills entered emotional intelligence communication skills presentation skills i went from a good career to a rocket ship career and that's why people that know me know that i am set with training soft skills and emotional intelligence and these skills they will build your cloud computing career your cloud architecture career your network architecture career faster than anything else in the world so i'm super excited to work with you if any of you guys have any questions on our training you can call our office somebody always asks us our phone this is our office number does anybody have any final questions for today any more questions for me questions on bgp questions i don't know if you have questions on router questions about cloud computing careers or questions about networking careers i'm here for you just post your question i'll do anything i can so let me know if you have any questions before we end the session we're happy to help any more questions for me i know there's delays so i'm going to wait one more minute to try and see if there's any questions otherwise tomorrow we're going to do it on zoom like we usually do but my associate my ceo chris has figured out how to live stream our zoom messages to youtube at noon eastern time tomorrow we are going to have how to get your first cloud architect job webinar and on this webinar we're going to teach you a lot of things kind of things that need to be on your resume the kind of things that hiring managers care about what is the actual job of a cloud architect so you know how to study it what is the things you need to know how do you interview properly and get hired and much much more we'll do that completely free chris probably has the link to that pin to the top i will ask chris from my team to actually paste it one more time so we'd love to see you in any of these free webinars everything we do is about building your career leo paradise thank you so much i remember when you first came to me as a student and i got to tell you the things that i've seen you doing the speed that i've seen you do it is amazing jeannie same thing with you um it's been absolutely absolutely absolutely wonderful having you as a student i remember when we spoke when you were in ethiopia for a while and now back then you're in us it's been a great experience um simon cordova no question for me but if you could please continue to post ahead of time absolutely so we try and post these uh available ahead of time you don't want to do too many because otherwise you'll be seeing months and months worth of live streams ahead of time and it gets confusing to people but we absolutely love it so we will keep posting them we are going to be doing a lot more and thank you so much for your kind words derek i am going to enjoy the evening with my wife um 25 years yes you know i don't even know what life was like prior to my wife but i don't think it was too good um nick love thank you so much i will definitely have a nice dinner when my wife comes back from her gymnastics lesson i've got a whole day planned around taking special care of her because she puts up with me and that's his alone not every husband has 10 servers floating around the house and routers and switches and firewalls in every room things that sound like jet engines so she puts up with a lot so i am really lucky for that so thank you so much any last questions otherwise i will see you all as very soon thank you reginald um alonzo always thank you so much i appreciate that more than you know um thank you for noticing our hard work and thank you for all your help alonzo with all the things that you do as well so i if there's no more questions i will turn this all back over to you enjoy the rest of your day if you found this content please share it with someone useful someone else who found it useful we love shares of our content if you have the ability to send it to others please do so if you're not a youtube subscriber please subscribe and hit the bells you'll be informed of these things sometimes we just go live immediately on the dime because we're inspired to do so but other times we do this so please join um thank you leo varieties it is an honor and a privilege to be part of your educational process whether it's free or paid we do not take that lightly it is very serious everything you do in your career we work very hard to give you the best free and paid content out there and if you'd like us to do any more if there's anything you're looking for send an email to askcloudmike gmail.com and i'll run it press my production team john thank you so much um it's been a pleasure since i met you very recently but it's all been a wonderful experience thank you all everyone take care and have a wonderful evening

Channel: Go Cloud Architects

Views: 8,647

Keywords: networking for cloud computing, networking and cloud computing, cloud computing technical skills, networking skills training, cloud architect skills, cloud architect career tips, cloud architect, cloud career training, cloud as a career, cloud career, aws networking training, classful and classless addressing, what is cidr, subnetting vlsm, vlsm exercises, subnetting, vlsm, cisco network training, aws bgp, what is bgp, aws networking tutorial, cloud architect training

Id: JmHWam2jKvo

Length: 238min 40sec (14320 seconds)

Published: Thu Aug 19 2021