FR and broadcast command

Guys hopefully i will get a good techinical answer to my question...i was going through FR mapping (static one with no inverse arp). In last we give a command broadcast.....FR is a NBMA.......environment and i have read the broadcast command is psudo broadcast what does that mean....can someone explain that to me plz

Thanks and regards

 

 

Comments

  • Unlike Ethernet, Token Ring and FDDI, by natural, FR is not a broadcast media. However, you could "simulate" the broadcast nature by   manually map layer 2 to layer 3 using the dlci


    On Tue, Aug 23, 2011 at 9:03 AM, tiger_woods <[email protected]> wrote:

    Guys hopefully i will get a good techinical answer to my question...i was going through FR mapping (static one with no inverse arp). In last we give a command broadcast.....FR is a NBMA.......environment and i have read the broadcast command is psudo broadcast what does that mean....can someone explain that to me plz

    Thanks and regards

     

     





    INE - The Industry Leader in CCIE Preparation

    http://www.INE.com



    Subscription information may be found at:

    http://www.ieoc.com/forums/ForumSubscriptions.aspx

  • Unlike Ethernet, Token Ring and FDDI, by natural, FR is not a broadcast media. However, you could "simulate" the broadcast nature by   manually map layer 2 to layer 3 using the dlci numbers. This is what they called psudo broadcast, meaning multiple uni-casts are actually in place. 


    On Tue, Aug 23, 2011 at 9:03 AM, tiger_woods <[email protected]> wrote:

    Guys hopefully i will get a good techinical answer to my question...i was going through FR mapping (static one with no inverse arp). In last we give a command broadcast.....FR is a NBMA.......environment and i have read the broadcast command is psudo broadcast what does that mean....can someone explain that to me plz

    Thanks and regards

     

     





    INE - The Industry Leader in CCIE Preparation

    http://www.INE.com



    Subscription information may be found at:

    http://www.ieoc.com/forums/ForumSubscriptions.aspx

  • FR is NBMA network and by default it doesn't carry any broadcast or multicast type packets. We can enable pseudo broadcast with DLCI mapping, which enables the to carry the broadcast or multicast packets over FR network.

     

  • Also, pseudo-broadcast use a special interface priority queue called "broadcast queue". This queue has its own buffers and can be tweaked according to the performance requirements you need to implement.

    Basically, the "broadcast" keyword enable broadcast and multicast traffic to be sent over the DLCI configured statically with a static map or dynamically mapped DLCIs using inverse-arp.

  • You know how what regular broadcast is and how/why it reaches all clients right? Considering it is a layer 2 broadcast, destination MAC address is FFFF.FFFF.FFFF. The router just send the oacket out the interface and the switch takes care to actually broadcast this packet out all its active interfaces except the one received upon. So in Ethernet, the infrastructure takes care of sending broadcast packets to all receivers, due to Ethernet nature.

    In Frame-relay, which is a NBMA media (non-broadcast multiple access), when a router wants to send a broadcast out a frame relay enabled interface, it puts the packet in its special broadcast queue. From here the packet is CPU processes and send outbound ONLY on DLCIs which are broadcast enabled; so basically the router sends multiple copies of the same packet out to broadcast enabled DLCIs cause the frame relay infrastructure is not able to perform the same function as an ethetnet switch (DLCIs are manually provisioned in the FR cloud). This is a pseudo-broadcast, process of router making copies of the same packet and delivering it out to those broadcast enabled DLCIs.

    Hope it is a bit more clear. Good luck with your studies!

  • See this example:

    I have IP/DLCI mapping without broadcast option

    interface Serial0/0.102 multipoint
     ip address 192.168.0.1 255.255.255.0
     frame-relay map ip 192.168.0.3 102

     frame-relay map ip 192.168.0.1 102

    I didn't get any reply while pingin on broadcast address

    R1#ping 192.168.0.255


    Type escape sequence to abort.
    Sending 5, 100-byte ICMP Echos to 192.168.0.255, timeout is 2 seconds:

    After that I added broadcast option

    interface Serial0/0.102 multipoint
     ip address 192.168.0.1 255.255.255.0
     ip pim dense-mode
     frame-relay map ip 192.168.0.1 102 broadcast
     frame-relay map ip 192.168.0.3 102 broadcast

    After this I tried toping on broadcast address:

    R1#ping 192.168.0.255

    Type escape sequence to abort.
    Sending 5, 100-byte ICMP Echos to 192.168.0.255, timeout is 2 seconds:

    Reply to request 0 from 192.168.0.3, 124 ms
    Reply to request 0 from 192.168.0.3, 124 ms
    Reply to request 1 from 192.168.0.3, 160 ms
    Reply to request 1 from 192.168.0.3, 160 ms
    Reply to request 2 from 192.168.0.3, 144 ms

    It is working fine and I tried to ping my multicast address as well

    R3#ping 225.1.1.11

    Type escape sequence to abort.
    Sending 1, 100-byte ICMP Echos to 225.1.1.11, timeout is 2 seconds:

    Reply to request 0 from 192.168.0.1, 200 ms
    Reply to request 0 from 192.168.0.1, 240 ms
    Reply to request 0 from 192.168.0.1, 228 ms
    Reply to request 0 from 192.168.0.1, 200 ms

    All working fine. So Peseudo broadcast forwards the broadcast/multicast packets over FR(NBMA) network.

    HAPPY STUDY!!!

    [:D]

  • Thanks giys for lovely answers....v nice of you......my question is in ethernet (broadcast domain) the pc broadcast because they need a mac address so in FR they already have dlci (which is layer two) then what is the need to do boradcast for...??? thanks

     

     

  • A host doesn't broadcast if it doesn't need to. To obtain L3 to L2 resolution (get a mac address) ARP is used - in the case of FR, ARP cannot be used so you can either statically map DLCIs and use the broadcast keyword or use inverse-arp.

    In Frame Relay the broadcast key is used also for multicast traffic. Broadcasts and multicasts are mostly used by routing protocols like EIGRP, RIP, OSPF.

    The use of a broadcast enabled DLCI is not mandatory per-se. For instance OSPF uses interface type NONBROADCAST in NBMA networks like frame relay by default. Other IGPs can be tweaked to use unicast updates. It really depends on how you want (or someone else wants) the network to behave.

  • Very well explained soul......my last question is why would you use static mapping (DLCI) over inverser arp ....Thanks alot again

  • It depends on the environment. The first thing that comes in mind about using static mappings over inverse-arp is PPPoFR.

  • Another reason for using static mapping over inverse-arp would be end-host reachability between spokes in hub and spoke topologies.

    Inverse-arp is unable to work over multiple PVCs therefore if the hub reaches the spokes via multiple PVCs there will be only a mapping to the hub on the spokes. To fix this you can either run a routing protocol or static map the spokes each other.

    Moreover, inverse-arp does not support IPv6 at the moment and the only solution to obtain an IPv6 connectivity over FR is to use static mappings.

  • From a time-management standpoint, static mappings will come up much quicker than dynamic mappings. Dynamic inverse-arp can drive you crazy while waiting for the protocol to come up - sometimes you want to "shut" and "no shut", but that just starts the wait again. Better to start the process and go off to do something else while inverse-arp takes it's sweet time.

    Static mappings will also prevent the router from sending inverse-arp requests over that DLCI.

    Very well explained soul......my last question is why would you use static mapping (DLCI) over inverser arp ....Thanks alot again

     

  • I was going through INE class in demand and in FR chapter (module 1) it sayd that Multipoint interfaces requires layer 3 to layer 2 resolution.???

    Eg which dlci do i use to reach ip address A.B.C.D....

     

    what abt point to point......wwhy doesnt it require mapping....i am really confusing it with ethernet in ethernet as layer 2 .....it alwats need MAC address to communicate so FR is layer 2 it will always require layer 2 DLCI so why there is no ,apping required (dynamic or static in point to pint)

     

    Thanks

  • what abt point to point......wwhy doesnt it require mapping....i am really confusing it with ethernet in ethernet as layer 2 .....it alwats need MAC address to communicate so FR is layer 2 it will always require layer 2 DLCI so why there is no ,apping required (dynamic or static in point to pint)

    Point-to-Point Frame Relay interfaces, such as point-to-point subinterfaces do not require layer 3 to layer 2 resolution because there is only one possible layer 2 destination reachable out the interface.

  • Hello tiger_woods,

     

    Addresses of remote destinations for each
    point-to-point subinterface are not resolved since they are
    point-to-point and traffic must be sent to the peer at the other end that's why we don't need mapping on point-to-point interface.

    The frame-relay map command is not available on point-to-point sub-interface only interface-dlci command is availble.

    [:D]

Sign In or Register to comment.