It is absolutely vital to keep in mind that the subject of this topic is by no means a service. An MPLS network or Multiprotocol Label Switching is actually a technique. You can utilize it to provide Metro Ethernet, IP VPN (Virtual Private Network), or direct it towards supplying optical services.
What this means is that an entire communication backbone may be built on this technique. The services which the users purchase, however, can NOT be called MPLS as such. You should choose to give the service an entirely different name. There is no real name set in stone, so the naming of your MPLS set- up is up to what the marketing department decides to call it.
The most essential concept of this technique is packet labeling. The way it is done in this type of environment, is fundamentally different from that of a normal IP set- up. The labeling in this instance makes the packet delivery not only faster, but more efficient as well.
In the case of the more traditional technique of direct IP, every router has to make a decision as far as where the packet of information should go. This decision had to be based entirely on the header section of a packet. The downside of this is that when a packet arrives, the router has to do some thinking, as to where to send it, which slows the entire process down.
When using the MPLS network technique, the packets' addressing is made that bit more efficient. This is done by assigning it to a certain FEC, which is indicated on the packet's label. Tables are provided to each router, showing them what to do with the different FEC packets. FEC stands for Forward Equivalence Class.
The upshot of this is that routers do not have to analyze the packet's title. All they have to do is use the label, to provide an index for the table. This is like reading a map book with an index for the router and the table, in turn, supplies the packet's new FEC.
The result of this is being able to give an MPLS network a "talent" for handling packets that meet certain predefined requirements, in a faster, more consistent way. It allows easier assigning of paths with lower latency (faster lines) to packets carrying real- time data. Doing this in direct IP is rather challenging, which has always been a major shortcoming of traditional routing.
Recently a lot of debate has gone on, about whether this service is part of Layer 2 or 3 of the OSI (Open System Interconnection). The major reason for the confusion is that the MPLS network technique can't logically fit into one of the seven OSI model's layers. In fact, one of the major advantages of this is being able to separate the mechanisms for forwarding, from the principal data linking service.
Further adding to the confusion, is that the two most widely used implementations of this technique are on two different OSI layers. The first of them is the Border Gateway Protocol, which is a Layer 3 service. The second of the most popular implementations is the Layer 2 pseudo wire VPN.
By Jacob Richard
What this means is that an entire communication backbone may be built on this technique. The services which the users purchase, however, can NOT be called MPLS as such. You should choose to give the service an entirely different name. There is no real name set in stone, so the naming of your MPLS set- up is up to what the marketing department decides to call it.
The most essential concept of this technique is packet labeling. The way it is done in this type of environment, is fundamentally different from that of a normal IP set- up. The labeling in this instance makes the packet delivery not only faster, but more efficient as well.
In the case of the more traditional technique of direct IP, every router has to make a decision as far as where the packet of information should go. This decision had to be based entirely on the header section of a packet. The downside of this is that when a packet arrives, the router has to do some thinking, as to where to send it, which slows the entire process down.
When using the MPLS network technique, the packets' addressing is made that bit more efficient. This is done by assigning it to a certain FEC, which is indicated on the packet's label. Tables are provided to each router, showing them what to do with the different FEC packets. FEC stands for Forward Equivalence Class.
The upshot of this is that routers do not have to analyze the packet's title. All they have to do is use the label, to provide an index for the table. This is like reading a map book with an index for the router and the table, in turn, supplies the packet's new FEC.
The result of this is being able to give an MPLS network a "talent" for handling packets that meet certain predefined requirements, in a faster, more consistent way. It allows easier assigning of paths with lower latency (faster lines) to packets carrying real- time data. Doing this in direct IP is rather challenging, which has always been a major shortcoming of traditional routing.
Recently a lot of debate has gone on, about whether this service is part of Layer 2 or 3 of the OSI (Open System Interconnection). The major reason for the confusion is that the MPLS network technique can't logically fit into one of the seven OSI model's layers. In fact, one of the major advantages of this is being able to separate the mechanisms for forwarding, from the principal data linking service.
Further adding to the confusion, is that the two most widely used implementations of this technique are on two different OSI layers. The first of them is the Border Gateway Protocol, which is a Layer 3 service. The second of the most popular implementations is the Layer 2 pseudo wire VPN.
By Jacob Richard
0 comments:
Post a Comment