2-10 Frame Relay Switching

  • A router can perform Frame Relay switch functions.

  • Both serial (major interfaces only) and ISDN PRI/BRI interfaces (Frame Relay over ISDN) are supported.

  • An interface can act as either a DTE, DCE, or NNI (network-to-network interface) device.

  • Traffic shaping can be used to control the transmission of outgoing traffic (out of the switch).

  • Traffic policing can be used on switched PVCs to control the behavior of incoming traffic (into the switch).

  • Congestion management can be used to specify how the Frame Relay switch informs other devices of congestion and how it reacts to congestion itself.

Configuration

  1. Enable Frame Relay switching:

     (global)  frame-relay switching  
  2. Configure the interfaces.

    1. Enable Frame Relay encapsulation:

       (interface)  encapsulation frame-relay  {  cisco   ietf  } 

      Only major interfaces are supported for Frame Relay switching.

    2. Configure the LMI type:

       (interface)  frame-relay lmi-type  {  ansi   cisco   q933a  } 

      The LMI type should be configured, because the router will be acting as a switch and will be providing LMI to other devices. The types are ansi (T1.617 Annex D), cisco, and q933a (ITU-T Q.933 Annex A).

      NOTE

      By default, LMI autosensing is enabled. For Frame Relay switching, the local router should provide one type of LMI to other devices. Therefore, you should configure a specific LMI type.

    3. Configure interface types:

       (interface)  frame-relay intf-type  {  dce   dte   nni  } 

      By default, a Frame Relay interface operates as a DTE (Data Terminal Equipment, dte ). The other routers and devices connected to this Frame Relay switch will likely be DTE devices. Therefore, the Frame Relay switch interface should become a DCE (Data Communications Equipment, dce ). If the router performing switching needs to connect to another Frame Relay switch within the frame cloud, the interface should become an NNI ( nni ) type.

  3. Configure the PVCs that will be switched.

    1. Configure static switching with no traffic control:

       (interface)  frame-relay route   in-dlci out-interface out-dlci  

      The PVC is defined from the incoming DLCI (in-dlci) on the interface to the outgoing interface (out-interface) and the outgoing DLCI (out-dlci).

      -OR-

    2. Configure switching for traffic control or ISDN:

       (global)  connect   name interface1 dlci1 interface2 dlci2  

      The PVC is given a text-string name and is defined from one endpoint (interface1 and dlci1) to another endpoint (interface2 and dlci2).

  4. (Optional) Use traffic shaping to control outgoing PVCs.

    Frame Relay traffic shaping can be used to characterize the parameters of a switched PVC. Traffic shaping is typically used when a router is acting as a Frame Relay switch to aggregate or concentrate multiple Frame Relay PVCs before sending a single data stream into another frame cloud. See Section 12-4 for further details.

  5. (Optional) Use traffic policing to control incoming PVCs.

    1. Enable policing on an interface:

       (interface)  frame-relay policing  
    2. Use a map class to apply policing parameters.

      • Configure the map class name:

         (global)  map-class frame-relay   map-class-name  
      • Configure the incoming CIR:

         (map-class)  frame-relay cir in   bps  

        The CIR is given as bps bits per second (the default is 56000). This is the CIR that the switch provides to devices on the PVC.

      • Configure the committed burst size , Bc:

         (map-class)  frame-relay bc in   bits  

        The committed burst size is given as bits (the default is 7000), based on the sampling interval Tc that is defined in Step 4). The Bc value is actually Tc multiplied by the CIR.

      • Configure the excess burst size, Be:

         (map-class)  frame-relay be in   bits  

        The excess burst size is given as bits (the default is 7000). It is also based on the sampling interval Tc.

      • Configure the 0 CIR measurement interval, Tc:

         (map-class)  frame-relay tc   milliseconds  

        PVCs can be provided with a CIR of 0 bits per second, offering no guaranteed throughput. When the CIR is 0, Tc defines the time interval when the incoming traffic rate can't exceed Bc plus Be. Tc is given in milliseconds (10 to 10000; the default is 1000 ms).

    3. Apply the map class to an interface or PVC.

      • Apply the map class to all PVCs on an interface:

         (interface)  frame-relay class   map-class-name  

        The map class is used as a template for all PVCs on a major interface. The map class can be overridden by other map classes applied to specific PVCs on the interface.

      • Apply the map class to a single PVC:

         (interface)  frame-relay interface-dlci   dlci   switched class   map-class-name  

        A single DLCI is defined on an interface. The switched keyword must be present when a map class is to be applied, along with the class keyword and the map class name.

  6. (Optional) Use congestion management to react to traffic congestion.

    1. Use congestion management for all PVCs on an interface.

      • Enable congestion management:

         (interface)  frame-relay congestion management  
      • Configure the discard threshold:

         (fr-congestion)  threshold de   percentage  

        When the interface's output queue reaches percentage of its maximum size, frames marked as discard-eligible are discarded.

      • Configure the congestion notification threshold:

         (fr-congestion)  threshold ecn  {  bc   be  }  percentage  

        The threshold for triggering the Explicit Congestion Notification (ECN) bits can be configured for the committed ( bc ) and excess ( be ) traffic rates. When the output queue reaches percentage of its maximum size, the ECN bits (FECN and BECN) are set in switched packets. The Bc threshold should be set equal to or less than 100%, and the Be threshold should be set equal to or less than Bc. By default, both Bc and Be are set to 100%.

    2. Use congestion management for single PVCs through traffic shaping.

      • Define a map class:

         (global)  map-class frame-relay   map-class-name  
      • Configure the discard threshold:

         (map-class)  frame-relay congestion threshold de   percentage  

        The threshold for beginning to discard frames is set to the percentage of the maximum queue size (the default is 100%).

      • Configure the congestion notification threshold:

         (map-class)  frame-relay congestion threshold ecn   percentage  

        The threshold for beginning to set the ECN bits (FECN and BECN) is set to the percentage of the maximum queue size (the default is 100%).

      • Set the size of a traffic-shaping queue:

         (map-class)  frame-relay holdq   queue-size  

        The maximum size of the shaping queue is set to queue-size number of packets (1 to 512; the default is 40).

      • Apply the map class to an interface or PVC.

        To apply the map class to all PVCs on an interface, use the following command:

         (interface)  frame-relay class   map-class-name  

        The map class is used as a template for all PVCs on a major interface. The map class can be overridden by other map classes applied to specific PVCs on the interface.

        To apply the map class to only a single PVC, use the following command:

         (interface)  frame-relay interface-dlci   dlci   switched class   map-class-name  

        A single DLCI is defined on an interface. The switched keyword must be present when a map class is to be applied, along with the class keyword and the map class name.

Example

A router is configured as a Frame Relay switch between two remote locations and another Frame Relay switch. The connection to remote location A is provided over interface serial 0/0, to remote location B over serial 0/1, and to the other switch over serial 0/2. DLCIs 100 and 101 at location A are switched to DLCIs 200 and 201 at the remote switch. DLCI 102 at location B is switched to DLCI 202 at the remote switch, and DLCI 103 at location A is switched to DLCI 103 at location B.

The connections to locations A and B are both configured for Cisco LMI. Serial 0/0 to location A is configured for congestion management (using FECN and BECN), a frame discard threshold of 50% of the queue size, and traffic policing using map class locationA for a CIR of 128 kbps and a Bc of 144 kbps. Figure 2-5 shows a network diagram.

Figure 2-5. Network Diagram for the Frame Relay Switching Example

graphics/02fig05.gif

  frame-relay switching   connect vc1 serial0/0 100 serial0/2 200   connect vc2 serial0/0 101 serial0/2 201   connect vc3 serial0/1 102 serial0/2 202   connect vc4 serial0/0 103 serial0/1 103   map-class frame-relay locationA   frame-relay cir in 128000   frame-relay bc in 144000   interface serial 0/0   description Remote Location A   encapsulation frame-relay cisco   frame-relay lmi-type cisco   frame-relay intf-type dce   clock rate 1300000   frame-relay policing   frame-relay class locationA   frame-relay congestion-management   threshold de 50   interface serial 0/1   description Remote Location B   encapsulation frame-relay cisco   frame-relay lmi-type cisco   frame-relay intf-type dce   clock rate 1300000   interface 0/2   description NNI to Frame Relay switch   encapsulation frame-relay cisco   frame-relay lmi-type cisco   frame-relay intf-type nni  


Cisco Field Manual[c] Router Configuration
Cisco Field Manual[c] Router Configuration
ISBN: 1587050242
EAN: N/A
Year: 2005
Pages: 185

flylib.com © 2008-2017.
If you may any questions please contact us: flylib@qtcs.net