68. Valant Encapsulation
The next important topic is We have VXLAN encapsulation. Now let’s understand that VXLAN Encapsulation means that you have your Inner Header means obviously the payload and the Inner Header where you have the excuse source and destination then we have the VXLAN Header, UDPHeader and then the Outer Header. Now, inside the VXLAN header you have we need of 24bits and that’s why we have possible 16 million segments. We have some reserve fields as well for future scope or purpose. Then you can see that we have a reserve, and one bit is on. That means that VX line is on or we need is on. So that’s the VX line header. Then you have the UDP header. You have the UDP source and destination. I’ll explain this in the upcoming slides.
And then you have the outer header. So that means that whenever the packet starts, it will get attached to the VXLAN header, the UDP header, and the outer header. and then it will go to the next stop. Correct. So suppose if you go and do the packet capture, you can see the packet capture will look like this. You have the actual source, and then you have the actual destination. Then once you have the actual source and destination you have the VXLAN header and here you can see that you have the flag and one of the bit icon that means the VXLAN network is true so conditions true so you have the inner header then you have the VXLAN then you should have the UDP youkan see that you have the UDP source that will be randomly generated number and the destination is 4789 incase of VXLAN in case of IBX land that’s used inside ACI will be the different port will see that. And then you have the outer header.
You can see the outer header is nothing but the VTEP. So suppose if you have two leaf these leafs are termed as a VTEP seven and VTEP two. And these VTIPS have IPs 1010, 51, and 1152. Correct. Then you should have some actual sources and actual destinations. Those actual sources and destinations have IP addresses like this. So once they will form the dynamic tunnel in between them. So I have two leaves, and then I am forming the dynamic tunnel in between them. So this is your VTEP, and this is your VTP, where we are doing the encapsulation of VXLAN and UDP. And then, from one location to another, having followed the available path, you are reaching out. Correct. Now what are the interesting fields or important fields we have inside this header format? Let me clean this up, so here you can see that inside the VX line header we have these fields. important one is the we need ID which is providing 16 million possible segments then about UDP header you can see clearly that source will be the hash.
So you can take any of the available paths to reach the destination. But the destination UDP destination port will be a fixed number. UDP 4789. Correct. So the source may be something that can take multiple paths. That’s why we have the ECMP load balancer with a fixed range of port numbers. And destination is one port fixed 4789 to reach from one place to another place. Then you can go and check the outer header. Source and destination will be the V tip addresses. And then you can check the automatic address as well. We have total 50 byte of overhead with the VX land header. Let me quickly go ahead, and you can see some of the explanation about those fields. So we have eight bytes for VXLAN divided into those. These four parts of three bytes are used as a VN ID to create the segments. Then we have the eight-byte UDP header of eight byte. The outer destination port is 4789. And in a destination in a source is the hash value in between. So here you can see the hash of L 2, L 3, and L 4 headers. So you can do the ECMP correctly. Then we have the outer IP header, and again, the outer IP header will be nothing but the source and destination IP of the V tips; apart from that, we have checked some UDP protocol and the IP header miscellaneous data, correct? So these information that whatever we have checked sofa these are belonging to the VXLAN encapsulation maybe Nexus OS also you can go and enable the Land and then this will work like this What about ACI VX Land information?
ACI VX Land is termed IEVX Land because if you have two different EPGs, EPG One and EPG Two, you need a contract between them, correct? So now if the VX land communication will happen, we have one extra field here for contract that will fall under flag. So we have SP, we have DP, and we have the PC tag as well that will tell you about the source EPG. And this flag means the context is applied or not.
So extra fields we have here which will tell that context is applied or not. The source EPG information is either there or not. And then there is a slight change in the UDP destination port. UDP destination port is 48879. Here you can see 48879, and this will be used inside the ACI. Although the entire IVX Land configuration is optimized, we don’t need to write a single line of code for this IVX Land inside the ACI. What we need to do is simply bring up the fabric, connect the endpoints, create the rules policy, and then attach to the leaf switches what we want from source to destination. What policy? What is the VM policy? What else? Four L seven integration et cetera et cetera and then the traffic will flow we don’t need to worry about the IVX Land configuration because this is fully 100% optimized inside the ACI fabric.
69. ACI Overlay Valant _ TEP 01
So here you can see that we have list or we have the agenda that we want to learn about different type of concepts one by one. So for inner capsules. example the endpoint learning the packet forwarding, the JDRF forwarding, a spine proxy are clean and then finally the forwarding software architecture and the ASIC generation because the magic behind the scene is happening inside the Asics how that this.
So we have leaves. different leafs. two one, leaf two and leaf three. And in all these leafs, you can see that first of all, they are part of same VRF, but we have different different bliss domain. So in leaf one, I have birdbath in leafing one, but leaf two, we have two. one and BD two. Now this PD at this point of time you can think bridge domain, you can think as similar to Landan then finally inside the leaf three you have the bridge domain two and you have the outside connectivity that’s the L three out we have the van connectivity. Now we have a different type of scenarios that we are going to discuss in this particular topology like within the leaf how the traffic will move across the leaf, how the traffic will move so let me show you quickly that what are the topics that we are go here, you can So here youkan see that we hav1234. The source leaf knows the destination.
So all those scenarios we have here listed so let’s do one thing, let’s complete the scenario number one and two and then scenario three and four that’s this fine proxy and flood we’ll take in the upcoming session. So the scenario number one is that source leaf knows the destination on the same leaf so when you are doing the communication on the same leaf remember still if you are indifferent EPG endpoint group you need contract and these things also we know from the previous knowledge from the previous labs that yes you need contract, you need subject and filter you need to apply some sort of agreement in between different EPG but the packet forwarding will be very much that is happening in the normal routing and switching world because by end of day they are in the same VRF, same VLAN and they are part of the same leaf switch so this will be very much traditional routing or switching now the use cases will come we’ll see in the upcoming scenario so different use cases will come and then how the ACI is handling those type of use cases that we’ll see now the scenario number two here is that if the source lift knows the destination so although you know you want to do the communication between EP one one to EP 32 but you know the destination now behind the scene how the packet encapsulation will happen so here youkan see the original packet has the source IP, Ands obviously, IPad obvioubecome these wilheader, correct? Souder care these these are the inner PA header, plus the header so these are the inner capsules Now, first of all, the switch here, which is the leaf one, will learn the end point. Obviously, whenever you have the connection, you have the end point, so first of all, your leaf one will learn again how this learning process is happening. Behind the scenes, we’ll discuss more and more in upcoming sessions.
o first of all this leaf will learn about this endpoint but since you know the destination correct so this spine proxy another flooding thing will not come into the picture because those things already happen or maybe spine has told you that what are the destination IP, Mac, et cetera so you can reach it so in this case just focus on the packet formatting so here you have your inner header and then you have the VX land tag and VX land tag is telling that the VRF or Bevin we know that we have L two and L three Vet and then you have the outer so here you have I small I stand for inner and then small O stand for outer and then obviously you have the source back and destination back of the leaf switch since you know the destination you are not doing the query to the spine switch otherwise this can fall under spine proxy methodology so what will happen that this spine switch actually they will forward the packet just seeing the outer header they will not go deep inside the packet and check the inner header and then they reconstruct their packet and on behalf of leaf one they will send the packet to the Leaf number? Maybe two or three. Or maybe all the other Leafs, except from where they are getting the request correct. So in this case, the spines will simply forward the packet to its destination. And in the destination leaf, the recapitulation will happen, and again, the reverse traffic will follow the same path. Correct. So send it to the end point, and that’s it. So this was the scenario number one and two. Let’s just stop here. The next section will follow. We will cover the rest of the scenario.
70. ACI Overlay Valant _ TEP 02
Let us continue where we left off. In the last section, In this section, I’m going to COVID scenario number three and scenario number four. Spine proxy in the flood So let’s start with the spine proxy now. What is happening? In this case, that leaf doesn’t know the destination. So what this leaf will do is go and do the query with the spine. Correct. Obviously, you don’t know the route, but you know someone I can ask, and he will respond on my behalf of me.
So here you can see the inner header will be intact the same. We have the VXLAN encapsulation, but the destination IP is inside the outer header, which is any cast TEP tip. Correct. Now how this spine will behave obviously spine should also know what’s the destination. Correct. And again, obviously, there are some other development processes going on behind the scenes as well. So at this point in time, we are understanding the packet behavior. What is fine, we’ll do; you can see here that what is fine is that we’ll send one packet on behalf of leaf one. So here, you can see that spine is sending the packet on behalf of leaf. But the header inside the header you can see clearly that header is leaf one and destination is leaf two. Inside the source and the destination IP outer headers. Correct. Now, once this particular packet is opened because the destination is already defined in the destination IP and Mac address in this packet, So once this packet reaches the actual destination, the destination will understand that this packet belongs to me. And obviously, the decapitation will happen, and it will get that packet correct.
So here you can see in this particular slide, which is a summary type of slide, the request that is going to the spine and the response that is coming from the spine. And then obviously, after that, these lifts will form the dynamic IVX LAN tunnel. And next time, whenever the packet goes, the spine will make the decision only based on the outer header format. It will not go and open the entire packet rather than it will take the decision on the outer packet header correct. So this is the way that the spine proxy is working. Let’s quickly go and understand. That is how the flood is working. because flooding is not as simple as it looks. Flood simply means that if you don’t have a destination, then send that request or send that query packet to all the receivers. Correct? Or send to all whoever has the IP, whoever has the destination IP he will respond if you do the flood either it’s inside the VLAN but in this case we are going to do the flood inside the bridge domain but anyhow if you do the flood then what will happen? Is that not a good method in a data center when you are flooding the entire bridge domain? In this case, in ACI, this flood is even optimized. And this flood concept you’ll find now in upcoming two, three slides is that it’s very much like Fabric Path flood.
Now you are flooding the routed domain. So here you can see that we know that for IP reachability, the ISIS protocol is running, and then we have the Coupe database, which has the consistency mechanism, which means one spine is learning the endpoint information, and then it will be replicated among these fines. So that’s why you have a consistent database as well. And I suppose if you want to communicate with the external world, then you have the multi-protocol BGP. And then finally, the IVX land tunnel is forming dynamically whenever the data exchange occurs there from one tip to other tip. So these are the control protocol mechanism we have inside that we want to do the flood inside the bridge domain. Correct? Now, when we are doing the flooding inside the bridge domain, at that time we have this zipper, that group outer header, and inside that we have the multicast traffic. What is happening in this case is that they will go, and they will form the multi-destination tree. And in that multi-destination tree, they should go and select one of the routes.
Correct? So suppose in our case we have two spine. Here, you can see that I have a spine one and a spine two. One is the route; one will be selected as a route, and he has the authority to build the multidestination tree with the help of the forwarding tag, or F tag. So that’s actually a very important concept because, although we are doing the flooding, this flooding is very well optimized, so we are not flooding everywhere. It’s just like we are not doing the flood each and everywhere, but we are flooding in group because we are creating the multicast destination tree. And the good thing about this FTAGID is that we’re building it here, and it’s very much like Fabric Path. So if you want to learn more about this, you can refer to Fabric Path, because Fabric Path is also under the routed domain. Whenever we are talking about routed domain routed under Lay at that time, automatically we are talking about ECMP, equal cost multiparting. So what is happening to get from one place to another place? If you have three link, since they are the routed fabric and in routing you have the equal metric or equal cost path. So at that time if you have three different flow, you can equally distribute these three different flow into three different paths.
Correct. And how many paths are supported? There are twelve different paths are supported inside this flood mechanism. So at the moment, you have the query. Suppose the leaf will send some query along that spine. I don’t know that this particular destination exists, and I need your help. Could you please tell me this destination now? Since this fine is working as a route for multi-destination paths, So what he will do is raking as a round once he will do the flood at that time suppose here you have the destination. The destination will respond. Obviously. That okay. I’m going to respond to you because I am the actual destination. But what about the other 100 leaves? Because that flood will go inside the same breeze domain. Say, for example, breeze domain one. And if that breeze domain is configured inside 100 leaves, all the rest of the leaves will be sent; say, for example, I don’t have the destination set for any of the nearby physical available spines. And then Spine will simply rule out this thing.
Okay, this can’t be the destination because I got it from the other spine. So something like mix and match type of thing that behind the scene is working just to provide the high efficiency and see good bandwidth utilization. Unlike that in SDP. Even you can’t send those data request data packets to the blocked links here. The agenda is that all links are equally utilized. Any link can form the VXLAN tunnel for the packet forwarding, and here you can see in the diagram as well that you are sending the request. Then flood packet is going everywhere and the nearby lift to the nearby is fine. We are responding that I don’t have this destination, and then the spine is sinking the database across different spine. Now, here you can see the packet format that the request is going to the spine whoever is the route. Maybe spine one and spine two. And then from there the flood packet is going everywhere. Flood based on the “if” tag and the “forwarding” tag. And then, once you get the response, you will know that this is the actual destination. And then the recapture will happen, and the final package will reach its actual destination.
71. Endpoint EPG EP Learning _ COOP
This is the important session where we are going to learn about the endpoint that we have already covered. But we are going to do the revision Edge also know about EPG that I have already covered then endpoint learning coupe and how everything fits in the same place obviously inside the ACF fabric. So let’s do a quick review of all these terms. Maybe a few of the terms are new for us in this particular section. Like what is the role of VLAN inside ACI.
Remember, we have VLAN, which means we can use VLAN as well as access VLAN, and we have other VLANs as well. The rest of the terms that we are seeing here, we are going to do a quick revision for that. So starting with the endpoint, what is an endpoint? Obviously, your server, any type of endpoint networking device, maybe fixed devices—those things can be your endpoint. Now what is happening that you leave switch? They are learning the endpoint entries, and then they are advertising those entries to the spine. So you know that you have your endpoint whatever endpoint you have mac and IP leaf is learning and then with the coupe message they have some coup synchronization message in between Leaf and spine they are sending those information to the spine and then spine is maintaining the database and syncing the database across the different spine. Correct. So what’s the important term here? like the legacy term related to rib. So here in ACI, you also have rib (without the slash) 32 because that’s 32 or maybe 120 18 IPV 632 regarding IPV 4 addresses.
So Mac and 32 they are termed as an endpoint. Correct. And ARP is used only for Letter three out. That means if you are going and talking-to the outside world at that time only your ARP will come into the picture. So these things have already been discussed in detail earlier. Now let’s move further and understand more about other terminology. So one of the important terms we have is the local station table, the global station table, and the others in the next slide. I’ll show you that local station table. At this point in time, you can simply refer to this as a traditional approach to learning logic. So what is happening in traditional networking means suppose if you have a switch and this switch, suppose this is 3850 switch, any of the end points connected with this or any of the server or PC laptop connected with this device. First of all, it will register the Mac address. If that device wants to communicate with any other device except this particular interface, the packet will be flooded inside the VLAN, et cetera. Correct. So that’s the traditional adjacent learning logic.
The next phase or next step to that is to use the convergent seasonal learning, and that will be used inside the VX land where you are forming the dynamic tunnel from one leaf to another, or from one TEP to another, correct? So that’s what conversational learning means: when you are doing the communication at that time, you are only learning those Mac addresses. So rather than maintaining the big amount of Mac table you have the shorter entries, shorter entry of the Mac entities. In our case, that’s obviously the end point. Then we have the external longest prefix mask, LPM. This LPM is used actually basically it issued for L three out type of communication.
So when you are communicating with the external world, at that time you have to build the routing table, and at that time the LPM table is being used. Finally, we have a fourth type of table, which is the proxy table, and this proxy table is maintained by this pin switch because, with the help of the coupe console of the Oracle protocol, they are maintaining the endpoint database and syncing those endpoint databases to the other pin switch. Remember that they are not advertising those endpoints to all the leaf switches because, again, if they do so, if they send the endpoint information to all the leaf switches, that will again become unnecessary. So unnecessary we are fulfilling the cam entry of the leaf switches rather than it should be stored at the level of the spine and when itis queried then it will respond. Correct? Now this is querying the spine, and then it is responding. It’s very much the logic of the hardware or spine proxy or hardware proxy. Again, I have one big flowchart. I will show you about layer two and layer three is fine hardware proxy later on. All right, so the next topic is the endpoint group. Endpoint group is again theological grouping of host endpoints. See here in the diagram that you may have say endpoint group EPG group, web, app, DB correct.
Now the web will not communicate to append DB because different endpoint group if you want to do the communication you have to have contact in between same endpoint inside or say different endpoint inside, same endpoint group. Suppose I have EPG called web and then I have endpoint say eleven 1213 they can communicate teach other there is no problem because the endpoint belonging to same endpoint group they will do the communication but different endpoint obviously they will not do the communication because this is the whitelist model. You need to have the ACL filters permission to do the communication. So that’s the end point the endpoint is used for again to group same type of endpoints. Correct. So that’s the use of endpoint groups. And then on the next slide, I’ll show that VLAN in the ACI. We’ll see in the next slide more detail about VLAN in ACI. But you can think at this point that this bridge domain is the actual VLAN. So whatever the VLAN rule we have in the traditional data center environment that role has been taken in much wider perspective in much wider range inside the bridge domain. What happens with the forwarding? We know that if you have the same EPG, there is no problem. Different EPGs obviously need the contract. Now, how can you verify this?
Here, you can see that you can go to the tin and you can go to the application profile. application profile, and then you can see the application EPG. So I have three EPGs (11213) inside one of the application profiles. This is the GUI way that you can see if you want to check via the CLA command. Again, we can go and check show endpoint VRF. That will show you the details about the end point. Now let’s quickly have a look at the VLAN. In the ACI, we have two types of VLANs. Actually you can think at this point of time three type of VLAN. One is an access VLAN that has local significance. So suppose if you have one leaf switch and then here you want to allow certain VLANs 1020.So you can make this an access VLAN. Correct? But remember, they have local significance. Their significance is only up to this point because at the moment you are coming to these points, I should draw only one. But okay, you may have multiple BX 9 tunnels. So at the moment you are here at this leaf point, and at the moment you are going inside the fabric. Suppose this color is the fabric.
The moment you go to the fabric, your VXLAN tunnel will come into the picture, and this local VLAN is again locally significant. It’s not significant inside the fabric because inside the fabric the things will happen with the outer header. Inner header will get encapsulated with help of outer header, VX land and outer header the packet will get forwarded. And again, when we are talking about the outer header and the VXLAN, remember, it’s very important. You have two terms. One, you have the BD, and then you have the VRF. So let me draw some other place. So you have two term, you have, this domain and you have VRF. If it is L-2, obviously you will use the BD. But if L three packet you will use the VRF and this VRF belonging to VXLAN L three and again VXLAN L two. So that means that in terms of VLAN, you know that in VXLAN we can have the 16 million we need, correct? We can have 16 million ideas that we can assign to the fabric. Correct? So again, this 16 million with respect to layers two and three—that’s the whole point. Now here in the diagram, you can see that you have the global VLAN that belongs to VXLAN. Then you have the SVI for the BD bridge domain, and then you have the access VLAN. Remember, access VLAN one. Another important VLAN we have is the platform-independent VLAN. This is a Pi VLAN that will be assigned by the ACI itself. Now how we can verify that, how you can see that.
So if you go and run this command to the leafs how endpoint IP and then the IP address, you can see that this is clearly the local Mac and the local IP. But here you can see the information about the access in cap VLAN assign locally to that particular interface. And then you can see the platform-independent VLAN as well. If you want to see more about platform independent wheel and then you have the command show VLAN ID 1719 extended. So here you can see the Pi, correct? And this pi VLAN. And then here you can see the access in cap VLAN and then you can see the VX land. We need information as well. So this command is important with this command—actually, you can see most types of VLANs in a single command, correct? Now next we have the important category of the endpoint types. What types of endpoints do we have? At the moment we know that at least two local and remote.
So local endpoint you can check with this command here you can see the access in cap VLAN, the local endpoint entry, the interface, you have a platform independent VLAN as well pitmen you may have the end point related to ABS and Ave. So if you have the application virtual switch, you can have the end point. And again you can see this is again a big topic is ABS NAV but you can see that you have a scope in the green color and then you have the endpoint here. This is a local learn where the tunnel and because VX line is there, so mostly you are seeing the tunnel information in the last column, correct? Then you have the remote endpoint. Again, at a remote end point, you are learning from somewhere, and that’s where you have the VX line information. You may have a Mac or an IP for the remote endpoint. Finally you have the endpoint related to VPC and that will be defined as a small O and very much they are termed as orphan port on VPC pier. So these are the category of the endpoint is four category of the endpoint. Mostly we you will get the use case related to local and the remote endpoint. Now, continuing with this fact,
72. Endpoint Learning
Let’s continue where we left off. So we have discussed different types of endpoints. Let’s quickly go over endpoint learning related to local and remote endpoints. Now, at this point in time, we know that different types of tables are there. So for ACI, we have a rip table, an endpoint table, and an ARP table; we know what the differences are between those tables and traditional. Now, what is happening, what type of packet of frames you will get and what you will do with those packets. So here you can see, first of all, that you can go and run this command, “show endpoint, IP,” and from the IP you will go and get the access in Cap VLAN. If you are seeing “L,” that means that’s the local, and you can see that that is your front panel port.
So suppose if you’re getting the frame in the front panel port, obviously you will learn the Mac address, correct? And if it is a routed packet, meaning if it is a layer 3 packet, then you will go and store the IP address, correct? So either it’s a routed packet or it’s our packet, where the switch will do the lookup. So in those cases, we’ll go ahead and store the packet as a L-3 packet or the IP information of that particular packet. Now, this is related to local endpoint learning. What about the remote endpoint? We know that in the case of a remote endpoint, either we have the Mac address or we have the IP address correct. Now here you can see that the CLI command that if you want to know detail about the remote endpoint, either Mac or IP, so you have to use this show endpoint and then Mac and IP, that’s the keyword. And then you can give the specific or exact Mac the IP address, and you get the detail. Obviously, this is the remote endpoint. So you will see that you are getting this information from a source other than the front panel port. That’s the interface; that’s the physical interface. But you’ll get this information from the spine, correct?
And that’s why you have this tunnel information because anyways, from two different leafs, leaf one and leaf two, you have any dynamic VX line tunnel, correct? Now, in this case, if it is a L two pack, obviously you will go and store the Mac address. If it is L three packet or it’s L three packet, that is again I’m learning with respect to my spine. Then I’ll go and store this entry as an IP entry. So mac Entry and IP Entry. Obviously Mac is the frame and the IP is faggot. So like that, we can differentiate. Now, here we have the important thing. So here you can see that we have Void, and if it is related to L2, that will again go to the bridge domain. Because we know that loosely bridge domains a VLAN, that’s a loose term. And now if this VNID is related to L3, obviously we have the VRF information, and that’s the reason here, as you can see in the output. So let me clean this up and highlight only these two points.
So if you have the Mac address, then you should have the BDVNID, correct? And if it is an IP address, at that time you have the VRF that belongs to that particular IP address. So you can go and understand that if it is the L two back and then it’s it is Two extended network, obviously why you will go and do the lookup for the IP rather than you will do the lookup only inside the segment, inside the bliss domain. If it is L 3 packets, then obviously it will get resolved, and then you will store those as a cache entry. Correct? Again, we have some important output here related to the tip TEP general endpoint, and obviously the key is the show endpoint. Then you can use either Mac or IP, and then you can learn about that Mac IP or that IP address. Then you will get the information related to Vnid. You will go and get the information about the tunnel that the VX land tunnel. You will go and get the information about the Pi.
Obviously, you have the VRF information as well. So those information you will go and get. If you are specific about the tunnel, you can go and check the tunnel. So from here I got the information about tunnel. Now this tunnel I want to see. So here you can go and check the tunnel information, where you’ll get what the tunnel is, what the source is, et cetera. And then again you can go and use this FN Read ACI dig Fnvread. And then you can go and verify which particular switch belongs to this particular TEP TenAl entry. So here actually these series of commands are in connection. First of all, you want to check your end point. So suppose I have my endpoint. So you can see here the endpoint entry learn over the leaf then which VX land tunnel you may have different type of V tap. So I have one VTP, but what is a dynamic tunnel? So these V tips of these step, which channel this particular endpoint entry, which particular tunnel dynamic tunnel it is using from source to destination if you want to know that information.
So first of all, just check which tunnel it is using, and then you can go and check the information about that particular tunnel. Now this particular tunnel belongs to which particular tap entry? So all the information you will get from this is correct. Now let’s quickly check the coup entry, how this coup is significant, and what it is doing at the moment. We know that at the moment the endpoint will get registered. So at the moment, the leaf will learn the endpoint entry. Then what will happen? That leaf will inform that particular information with the coupe database, actually with the coupe message to the coupe database spine, having the coup database and leaf using some sort of coupe hash method to inform their spine that this entry I have learned youkan store inside your database. Now, by default, in normal scenarios and in normal conditions, the spine will not do the route reflection. Once you learn all the Macaroni, they will not gonad push all the Macarons to all the leaf switches.
Because in that case, again you can understand that unnecessarily you are putting all the Mac information or all the endpoint information to all the switches rather than what is happening, that if you have any bounce entry at that time only that spine will go and teach that these are the Mac entities has been changed. Because I got to know this is the new updated entry in my database. And you should also know that. So maybe at this point of time, if you don’t know that, what does it mean by the bounce entry? So here you will see that normally this point does not push any database entries to the leaf. It does receive and store it. The exception are the bounds entries, correct? So what is happening? Suppose IPA and IPB are the source and destination, is the destination. But in any case, maybe due to Motion or any other technology, this particular b. So Mac b and IPB, you have the Mace and IPB, they move to this location. Now this leaf still knows my destination is here, and it will go and send the frames to this location. But it has been moved.
Now that it has been moved, the leaf will learn that endpoint, he will register that entry to the coupe database, the coup database will remove this entry, and then he will go and add that Mac IP B belongs to leaf 3. And in that case, he will tell him, “Let’s see, the destination you are looking for is not over leaf two, but it has been moved to leaf three.” That’s something termed a “bounds entry.” In case of bounds entry only the coupe database will teach the leaf or they will inform the leaf, which is that the entry has been changed and then your destination has been changed. Now that’s the one use, that’s one of the special use. You have other uses for this fine proxy. Suppose you want to send the frame or the packet to some destination you don’t know, then you will go and request. You will send some sort of query to the spine, and the spine will work on behalf of you. And then it will try to search the information, the destination, and the destination location information, and then it will tell you about this thing that we have actually already discussed. But these are the important terms related to as pine, related to coupe, and related to the coupe database.