Bgp which osi layer

All anyone can offer are opinions, because there is no organization that is in charge of defining what layer a protocol is at. I disagree with some of your statements. The OSI model is much more than a "logical way of describing features and functions of networking.

These all exist to create international standards for all coutries to follow to ensure inter-operability between countries. OSI Protocol Suite is not used anymore.

All that's left from OSI is the logic behind it, which yes manufactures may follow to allow interoperability between different features and functions of networking. But if they were hard-set standards which everyone followed, you wouldn't have any arguments finding out which OSI layer BGP belongs to.

OSI says that L3 is about path determination and routing, which makes it pretty clear to me that all routing protocols fall into L3. Doesn't really matter though, because again, it's all just logical representation. Where are you getting your facts?

Who says the OSI protocol suite is not used anymore? Do you know of anyone using the OSI protocol suite? I haven't heard of anyone using it in many years. I'd be surprised if there were still organizations using it in this century, honestly. About this question, I want to say, we should firstly think about what's the routing protocols?

We use routing protocols to carry routes, to build routing table, to update routing table. All of these functions are to find route for packets. From this point of view, a protocol is called routing protocol does not depend on which layer it operates, it depend on the function of it. If it can provides functions of routing protocol, it can be called routing protocol, no matter it operates which layer. The OSI model is just that, a model, and it really only applies directly to OSI protocols, which is not what we're talking about here.

You have to keep a couple of things in mind. First, there are two types of traffic we could be talking about: control plane and data plane. The OSI model, as taught by Cisco, really only applies well to data plane traffic, i.

The OSI model as taught by Cisco does not apply well to control plane traffic. Routing protocols are control plane traffic, not data plane traffic. It is a layer three control plane management protocol that happens to use TCP for reliable communication with neighbors. In the world of computer networking a local executable software process that performs computations that result in an output data structure is an application.

It may not be a "user" application but it is an application to the local system. When the destination host receives the datagram it deletes the "L3" header and, if applicable, passes the packet to the appropriate Transport layer protocol who, in turn, after validation and sequencing delivers the data stream to final application program for processing. The L3 operation is not performed by BGP but by the specific switching process that forwards the packets.

The switching process is performed based on the information contained within the RIB, which may have been built from information contained within the BGP information base.

Within the local operating systems process structure BGP is not operating at Layer 3. Good to see you back on CLN. I really like the explanation and point of view in your recent post. I second the motion on Douglas Comer's book. I'm not entirely sure what you're saying. Yes, you could arguably say that from one perspective BGP is an application layer protocol, but if I'm remembering correctly, that wouldn't be the case in OSI.

OSI makes the distinction between the data plane and the control plane. Routing protocols are control plane. They are management protocols for the network layer. In true OSI, routing protocols reside at the network layer. Your definition of application is technically correct, but is not the definition used by the ISO when defining the application layer. In OSI, the application layer included virtual terminal access, file transfer programs, directory services, etc.

These sorts of painful discussions are the result of Cisco trying to shoe-horn everything into a simplified version of the OSI model and it just doesn't work.

Don't even get me started on the "what layer is ARP" question. Actually, the answer to that question is similar. It operates at layer two, but it isn't really a layer two protocol.

It's actually a subnetwork convergence protocol designed to be the bridge between layers two and three and it directly carries information about both layers. Trying to make it fit perfectly into the simplified Cisco version of the OSI model is a fruitless endeavor.

If you first ask if the protocol is control plane or data plane and then choose the appropriate reference model you will find that most protocols fit rather easily. If you want to define the application layer that way then you're right. I just disagree about making up our own definitions of what constitutes an application. These conflicts, however, are more frequent when one only looks at the original OSI model, ISO , without looking at the annexes to this model e.

In response to your comment, "I just disagree about making up our own definitions of what constitutes an application. ErickB, who ever said that BGP was not a routing protocol? There is a general rule regarding hierarchical layers and that is, no matter what layer the protocol may start it can only go down the stack. It can not go up the stack and then back down. Do you recall your layers for each reference model?

Which jneiberger agreed to regarding "User Data" traffic in his post No one was trying to "squeeze" protocols into a particular layer.

We were trying to show hopefully you picked up on it that there are different ways to "view" how protocols are referenced to the different models available. As Kevin said, I am not going ot loose sleep over it and I do not want to argue either. Thank you for your posts. And how vain of you to think I would be addressing you? I don't recall stating that you said that did I? I did not accuse you have saying that said anything.

I guess your comment was just a bit out of touch with the current discussion and I did not recall someone else saying that BGP was not a routing protocol. I am vain, seems a bit harsh dude. Then again my comment about you guys reading the entire post could be considered equally as harsh. I apologize for that. While you may not be my friend in the sense of the word as I do not really know you. I do consider you a colleague and respect your views and opinions. Just because you and another member of the forum were having a discussion, doesn't mean I wanted to join in.

I simply read the OP question about why it's a routing protocol, so I answered. If you look, which obviously you didn't my comment was directed to the OP. That why I called you vain. You replied to something that was not directed to you and didn't take the time to see if my post was even directed to you. I apologize if you took the vain comment to heart. I wasn't using the term in a derogatory sense. I apologize if you took it that way. I said that because multiple people replied to that post over time and you somehow thought I was directing my post to you when I gave no indication os such.

Again, my apologies if you took that the wrong way. Presumptive would have been a better word, but I didn't think about that word yesterday.

I didn't mean to refer specifically to you when I mentioned making up our own definitions. What I mean is that the models we use have fairly specific definitions of what an application is. What we think of as an application these days is not quite what the OSI and DoD developers had in mind. If we want to use the model, we have to agree to use its definitions.

By defining any type of process as an application layer function, we've violated the model. The strict adherence to a vertical model is a side-effect of the way Cisco teaches these models and it does not relate directly to real-world protocol operation or development.

Active 4 years, 2 months ago. Viewed 12k times. Improve this question. Maha Maha 2 2 gold badges 4 4 silver badges 10 10 bronze badges. Add a comment. Active Oldest Votes. Improve this answer. Thats exactly what I concluded after reading many articles online.

You reaffirmed this fact. Xavier Nicollet Xavier Nicollet 3 3 silver badges 9 9 bronze badges. Sign up or log in Sign up using Google. Sign up using Facebook. BGP neighbors, called peers, are established by manual configuration among routers to create a TCP session on port A BGP speaker sends byte keep-alive messages every 60 seconds to maintain the connection. A route announcement is sometimes referred to as a 'prefix'.

A prefix is composed of a path of AS numbers, indicating which networks the packet must pass through, and the IP block that is being routed, so a BGP prefix would look something like: 42 It modifies the routing table to maintain the autonomous systems that are traversed in order to reach the destination system.

BGP attributes have different distinctive types that defines how routes are going to use and propagate a certain attribute to it's neighbors. If there are no specific settings that can affect the outcome, BGP Best Path Selection Algorithm determines the best route by selecting the shortest path to the destination. An Autonomous System is a single network or a set of networks and routers, which are under the control of one administrative entity.

Add Peer D, and set the AS number at the individual neighbor level. On which layer bgp works? Asked by: Franco Bahringer. Why is BGP used? What layer is TCP on? What layer is OSPF? What layer is SMTP? RIP works at layer 3 and sends routing information across the network.

What is a Layer 4 protocol? What are the BGP states? What is BGP port? What are the Layer 3 protocols?