CCIE Practical Studies, Volume I

Source Route Bridging (SRB)

At the same time that the IEEE 802.1 committee was considering adapting transparent bridging for the standard to connect LANs, it also was reviewing source-route bridging as an alternative. As history tells us, the IEEE 802.1 committee adopted transparent bridging. When this happened , source-route bridging was presented before the IEEE 802.5, where it found a home as the protocol to connect IBM Token Ring LANs and IEEE 802.5 LANs. Figure 13-8 represents the 802.5 Token Ring frame.

Source-Route Bridging Overview

Source-route bridging uses a combination of explorer packets and a RIF field to determine the best path through the bridged network. A source-route bridge uses the routing information field (RIF) in the IEEE 802.5 MAC header of a datagram, as in Figure 13-9, to determine which rings or Token Ring network segments the packet must transit. This is where the "route" in source route comes from.

The RIF is inserted into the MAC header immediately following the source address field in every frame by the source station. The destination station reverses the routing field to reach the originating station. Unlike transparent Spanning Tree bridging that requires time ”50 seconds to converge in the event of failures ”source-route bridging allows multiple active paths through the network, which provides an extremely efficient way to use alternate paths in the event of a failure. Most importantly, source-route bridging places the burden of transmitting frames on the end stations by allowing them to determine the best routes for frames across the network. The IBM Token Ring specifies a maximum number of 8 rings and 7 bridges in the RIF, while 802.5 specifies a maximum of 14 rings and 13 bridges. Figure 13-9 represents the 802.5 MAC Frame.

The ring in a Token Ring network is designated in the routing information identifier (RII) by a unique 12-bit ring number ranging from 1 to 4095. Each bridge between two Token Rings is designated by a unique 4-bit bridge number in the RIF. The valid bridge numbers are 1 to 15. Bridge numbers must be unique only between bridges that connect the same two Token Rings. If the RII is set to 0, there will be no RIF in the frame; if the RII is set to 1, a RIF will be included in the frame.

The RIF Field

The RIF is composed of 16 “bit routing control fields and routing descriptor fields. Figure 13-10 illustrates the basic RIF format.

Figure 13-11 illustrates the routing control format for the RIF, followed by the descriptions of each field.

• Shaded fields are reserved.

• type ” The explorer type is used, as follows :

  - 00: Specific routes explorer

  - 10: All rings, all-routes explorer

  - 11: All rings, spanning routes (limited broadcast)

• length ” This is the total length in bytes of the RIF.

• D ” This is the direction, indicated as follows:

  - 0: Interpret route left to right (forward)

  - 1: Interpret route right to left (reverse)

• largest ” This is the largest frame that can be handled by this route, as follows:

  - 000: 516 bytes (DDN 1822)

  - 001: 1500 bytes (Ethernet)

  - 010: 2052 bytes

  - 011: 4472 bytes (Token Ring and Cisco maximum)

  - 100: 8144 bytes (Token Bus)

  - 101: 11,407 bytes

  - 110: 17,800 bytes

  - 111: 65,535 (initial values)

Figure 13-12 describes the routing descriptor format of the RIF string. When you configure a static RIF or when a RIF is represented, it is represented in dotted hexadecimal.

• Ring number ” Unique decimal ring number within the bridged network.

• Bridge number ” Unique decimal bridge number between any bridges connecting the same two rings. A bridge number of 0 indicates that the RIF is terminating.

Figure 13-13 shows an SRB network. The RIF from station Alpha to station Bravo would read as follows:

0830.0012.002a.00b0

The 0830 is the 16-bit RC field, and 0012, 002a, and 00b0 are the three 16-bit RD fields. The first four bits, from left to right, state that the explorer type is 0, or a specific routes explorer. The 8 says that the entire RIF is 8 bytes in length. The D bit is set to 0, indicating that the RIF is read from left to right, or forward. The next three bits are set to 011, which sets the frame size to be 4472, the Cisco maximum. The RD fields break down rather easily:

RING1-BRIDGE2 = 0012 RING2-BRIDGE10 = 002a RING11-BRIDGE0 = 00b0

A bridge of 0 tells the SRB to terminate the RIF and that no more bridges follow the ring. For more information on configuring a static RIF, see the section entitled "Configuring a Static RIF."

A source-route bridge can determine whether to forward an explorer into a ring by looking at the RIF field. An explorer packet will not be forwarded to a ring where a duplicate ring-bridge-ring pattern already exists in the RIF.

The information in a RIF is derived from explorer frames generated by the source node. These explorer packets traverse the entire source-route bridge network, gathering information on the possible paths the source node might use to forward traffic. SRBs use three types of explorer frames:

• All-routes explorer or all-rings explorer ” This type of explorer is propagated from ring to ring from each SRB toward its destination. The destination station receives an all-routes explorer and forwards a directed, nonbroadcast frame back at the source that originated it.

• Specific routes explorer or local explorer ” This type of explorer is used by the end station to locate a specific station on a local ring. NetBIOS and SNA produce these types of explorers.

• Spanning explorer or limited routes explorer ” Spanning explorer can be propagated only if the Token Ring interface has source-bridge spanning enabled. Protocols such as NetBIOS require this type of explorer frame. When an end station receives a spanning explorer, it responds with an all-routes explorer frame sent toward the originating station.

Figures 13-14 and 13-15 illustrate how the all-routes explorers and spanning explorers operate .

Configuring Source-Route Bridging

Source-route bridging can be configured in three primary ways:

• Basic local SRB

• Multiport local SRB

• Remote source-route bridging (RSRB)

Configuring Basic Local Source-Route Bridging

Local SRB in its simplest form exists between two rings on a router. Figure 13-16 illustrates this type of configuration.

To configure this type of SRB, follow this two-step process:

Step 1. Enable the use of the RIF, if required, with the router interface command multiring all. The full syntax is as follows:

Router(config-if)# multiring { protocol-keyword all other } no multiring { protocol-keyword all other } The Cisco IOS Software allows you to also specify a protocol. This is specified by the argument protocol-keyword. This keyword allows for per-protocol specification of the interface's capability to append RIFs to routed protocols. When it is enabled for a protocol, the router will source packets that include information used by source-route bridges. The protocols supported and the keywords are as follows:

apollo ” Apollo Domain

appletalk ” AppleTalk Phases 1 and 2

clns ” ISO CLNS

decnet ” DECnet Phase IV

ip ” IP

novell ” Novell IPX

vines ” Banyan VINES

xns ” XNS

Two other keywords are used with the multiring command. The keyword all enables the RIF for all frames, and this is the recommend use. The keyword other enables the RIF for any routed frame not included in the previous list of supported protocols. The no multiring subcommand with the appropriate keyword disables the use of RIF information for the protocol specified.

Step 2. Configure SRB for the Token Ring interface. This is accomplished with the following interface command:

Router(config-if)#source-bridge local_ring bridge_number destination_ring

The configuration for SRB is shown in Example 13-11.

Example 13-11 Local SRB Configuration

interface TokenRing0 no ip address no ip directed-broadcast ring-speed 16 multiring all RIF enabled source-bridge 1 1 2 From ring 1 thru bridge 1 to ring 2 ! interface TokenRing1 no ip address no ip directed-broadcast ring-speed 16 multiring all RIF enabled source-bridge 2 1 1 From ring 2 thru bridge 1 to ring 1 !

Configuring Multiport Local Source-Route Bridging

The other type of SRB is needed when there are more than two Token Ring interfaces to bridge between. This type of configuration requires a virtual ring to be defined on the router. A virtual ring is just as the name describes, a virtual entity that connects two or more physical rings locally or remotely. A virtual ring also is referred to as a ring group . As you will see in the next section, "Configuring Remote Source-Route Bridging," a virtual ring can span an entire IP domain. For now, the virtual ring will be limited to the local router. Figure 13-17 contains an example of a three-port SRB. To configure SRB between rings 1, 2, and 10, you will need to configure a virtual ring. Then, you will source-bridge every real Token Ring to the virtual ring. Figure 13-18 illustrates conceptually how the network will look with the location of the virtual ring.

To configure this type of SRB, follow this four-step process:

Step 1. Define a virtual ring on the router. This is accomplished with this global router command:

Router(config)# source-bridge ring-group virtual_ring_number The virtual ring number can range from 1 to 4095.

Step 2. Enable the use of the RIF, if required, with the router interface command multiring all. The full syntax is as follows:

Router(config-if)# multiring { protocol-keyword all other } no multiring { protocol-keyword all other }

Step 3. Configure SRB for the Token Ring interface. This is accomplished with the following interface command:

Router(config-if)# source-bridge local_ring bridge_number virtual_ring

Step 4. (Optional) Enable Spanning Tree explorers. By doing so, you can reduce the number of explorers that transverse the network. NetBIOS and NetBEUI require Spanning Tree explorers to function properly. Cisco recommends enabling Spanning Tree explorers in complex multiprotocol networks. To enable them, use this interface command:

Router(config-if)# source-bridge spanning

Example 13-12 shows the configuration for local multiport SRB for the network in Figure 13-17.

Example 13-12 Multiport SRB Configuration

! source-bridge ring-group 100 Configure a virtual ring of 100 ! interface TokenRing0 no ip address no ip directed-broadcast ring-speed 16 multiring all RIF enabled source-bridge 1 2 100 From ring 1 thru bridge 2 to V-ring 100 ! interface TokenRing1 no ip address no ip directed-broadcast ring-speed 16 multiring all RIF enabled source-bridge 2 2 100 From ring 2 thru bridge 2 to V-ring 100 ! interface TokenRing2 no ip address no ip directed-broadcast ring-speed 16 multiring all RIF enabled source-bridge 10 2 100 From ring 10 thru bridge 2 to V-ring 100

Configuring Remote Source-Route Bridging

SRB also can be configured to span a single WAN serial interface or an entire IP domain. This type of configuration is called remote source-route bridging (RSRB), which involves defining a virtual ring to link the remote bridges. Figure 13-19 illustrates a Token Ring SRB network connected by a common Frame Relay network.

To configure RSRB, you need to define a common virtual ring to connect all the SRBs. The most logical spot for the virtual ring is the IP network, or the WAN, in this example. Figure 13-20 illustrates the RSRB with the virtual ring defined.

RSRB offers a variety of options to encapsulate SRB information over the IP network. The options depend primarily on the WAN interface. RSRB can be configured with four types of encapsulation: direct, FST, TCP, and Frame Relay. All four types of configurations are similar, but some require additional configuration. Table 13-1 presents an overview of the four encapsulation types, described in greater detail in the list following the table.

Table 13-1. RSRB Encapsulation Types and Requirements

• Direct ” Direct encapsulation is a quick way to encapsulate SRB frames over a single physical network between two routers. This command does not offer some of the advanced features of RSRB, such as local acknowledgment, but it is efficient. If direct encapsulation is used on a WAN interface, it must run HDLC as the data link protocol. Remember, direct encapsulation can be used only between two routers that are adjacent to each other.

• Frame Relay ” Frame Relay encapsulation uses RFC 1490 to directly encapsulate RSRB in the Frame Relay frame. Frame Relay encapsulation allows you to control RSRB on a per-PVC basis. In addition to the RSRB commands, a frame-relay map rsrb command is needed to map the rsrb to a DLCI number on multipoint interfaces. This type of encapsulation provides less overhead than TCP, but it does not offer some of the advanced features of TCP.

• TCP ” The TCP type of encapsulation provides the greatest benefits, but at a lower performance. Cisco recommends using this encapsulation when connecting Token Ring bridges across heterogeneous networks. TCP also provides load balancing and local acknowledgment.

• FST ” FST stands for Fast-Sequenced Transport. Although it consumes less overhead than TCP, it still is not as fast as direct encapsulation. FST can be used to connect more than one SRB, but it does not offer local acknowledgment. An FST RSRB also needs to have an FST peer name assigned.

To configure RSRB, follow this four-step process:

Step 1. Enable the use of the RIF, if required, with the router interface command multiring all.

Step 2. Enable the virtual ring with the source-bridge ring-group virtual_ring command.

Step 3. Configure SRB from the physical ring to the virtual ring.

Step 4. Determine the encapsulation type to use, and configure RSRB.

Direct encapsulation: Create a remote peer for each peer router and one for the local router. Use the following global router command:

- Router(config)# source-bridge remote-peer virtual_ring interface interface_name

Frame Relay encapsulation: Create a remote peer for each peer router and one for the local router. Use the following global router command:

Router(config)# source-bridge remote-peer virtual_ring frame-relay interface interface_name dlci_number [ lf largest_frame_size ]

- You also must add a frame-relay map statement for multipoint interfaces; use the following interface command:

Router(config-if)# frame-relay map rsrb dlci_number TCP encapsulation: Create a remote peer for each peer router and one for the local router. Use the following global router command:

Router(config)# source-bridge remote-peer virtual_ring tcp ip_address [ lf largest_frame_size ] [ local-ack ]

The IP address is the IP address of the remote router that you want to reach. You also must have IP connectivity to this address.

FST encapsulation: First, create a source-bridge FST peer name. This peer should be a loopback address or a local Token Ring interface. Use the following global command:

Router(config)# source-bridge fst-peername local_ip_address

- Next, create a remote peer for each peer router and one for the local router. Use the following global router command:

Router(config)# source-bridge remote-peer virtual_ring fst ip_address [ lf largest_frame_size ]

The IP address is the IP address of the remote router that you want to reach. You also must have IP connectivity to this address.

NOTE

We highly recommend using loopback addresses for the RSRB peers and DLSw peers. By pointing the peer to a loopback interface, you provide the peer with an interface that can be down only if the router is down or there is a problem with IP routing. A physical interface can go down, which will cause the peer to drop. This, in turn , could affect other interfaces on the remote router that might need to see the RSRB or DLSW traffic. By pointing the peer at a loopback interface, the peer will always remain up, thereby servicing all ports on the router independently of each other.

Determining Source-Route Bridge Status

You can view the status of the SRB by using the following commands:

show source-bridge show source-bridge interface

The show source-bridge command displays information about all the SRBs on the router. With this command, you should verify the entries that you made to configure SRB with the output. Ensure that ring and bridge numbers match what you configured. If other bridges are active, you will see receive and transmit counts increment. This command also displays the number and type of explorers received on the network. The command also display errors and drops on the SRB. For more detailed information about this and all source-route bridging commands, read the Cisco Press title Cisco IOS Bridging and IBM Networks Solutions. Example 13-13 lists the output of the show source-bridge command.

Example 13-13 Status of the SRB

srb_router# show source-bridge Local Interfaces: receive transmit srn bn trn r p s n max hops cnt cnt drops To0 1 1 100 * b 7 7 7 7297 2 154 To1 2 1 1 b 7 7 7 2 390 0 Global RSRB Parameters: TCP Queue Length maximum: 100 Ring Group 100: virtual ring Number No TCP peername set, TCP transport disabled Maximum output TCP queue length, per peer: 100 Rings: bn: 1 rn: 1 local ma: 4007.781a.e789 TokenRing0 fwd: 0 Explorers: ------- input ------- ------- output ------- spanning all-rings total spanning all-rings total To0 0 6856 6856 0 1 1 To1 0 1 1 0 390 390 Explorer fastswitching enabled Local switched: 6300 flushed 0 max Bps 38400 rings inputs bursts throttles output drops To0 6300 0 0 0 To1 0 0 0 0 srb_router#

The show source-bridge interface command displays a quick overview of the SRB on the network. It displays the state, ring, and bridge number, as well as packets in and out. Use this command for a quick view on whether the bridge is operating and passing data. Example 13-14 lists the output of this command.

Example 13-14 Status of the SRB Interfaces

srb srb_router# show source-bridge interfaces v p s n r Packets Interface St MAC-Address srn bn trn r x p b c IP-Address In Out To0 up 0007.781a.e789 1 1 100 * b F 35373 7393 To1 up 0007.781a.e729 2 1 1 b F 30158 7228 srb_router#

Practical Example: Configuring Remote Source-Route Bridging

Figure 13-21 models three Token Ring routers connected by a Frame Relay network. To configure RSRB on this network, you can follow the four-step process defined earlier. The first step is to enable the RIF field with the multiring all interface command. You will perform this on all the routers in the model. Second, you need to define a virtual ring. The logical location for the virtual ring in this model is the Frame Relay cloud. Therefore, you will make virtual ring 100 reside over the Frame Relay cloud. To configure the virtual ring 100, use the global router command source-bring ring-group 100.

The third step involves configuring SRB on all the routers that you want to connect through the RSRB. Example 13-15 lists the SRB configuration of the shuttle_15 and shuttle_3 routers.

Example 13-15 SRB Configuration of shuttle_15 and shuttle_3 Routers

hostname shuttle_15 ! <<<text omitted>>> ! interface TokenRing0 ip address 172.16.1.1 255.255.255.0 no ip directed-broadcast no ip route-cache no ip mroute-cache ring-speed 16 multiring all RIF enabled source-bridge 15 10 100 SRB from ring 15 through bridge 10 to V-Ring 100 source-bridge spanning Enable support for NETBios and NetBEUI ! _______________________________________________________________________ hostname shuttle_3 ! interface TokenRing0 ip address 172.16.3.3 255.255.255.0 ring-speed 16 multiring all RIF enabled source-bridge 3 3 100 SRB from ring 3 through bridge 3 to V-Ring 100 source-bridge spanning Enable support for NETBios and NetBEUI !

NOTE

Using Windows 9 x NetBEUI to Test RSRB

In this model, you can test the bridge with Windows 9 x networking, or, specifically , NetBEUI. Without actual traffic, the RSRB will stay in a closed state. When the RSRB detects traffic, the bridge will transition from a closed state to an open state. In a lab environment, you can generate the traffic needed for the RSRB with Windows 9 x configured using NetBEUI for file and print sharing. In a Windows 9 x configuration, when the network neighborhood is browsed (or select Start, Find, Find Computer ), this forces traffic across the RSRB. When the RSRB detects traffic, the state of the RSRB transitions from closed to up. This is an effective and easy way to test the RSRB. Use this same testing technique to test your DLSw configurations. With two workstations configured, you can print and share files over transparent, source-route, or DLSw networks. Windows 2000, Me, and NT also can be used for testing.

In this model, you will be testing the RSRB with Windows 9 x networking. Therefore, you also will need to configure Spanning Tree explorers on the source-route bridge, as noted in Example 13-15.

The final step is to select an encapsulation to use on the RSRB and to configure it. In this model, we are using TCP and the RSRB encapsulation. Each router in the RSRB group will be configured with three source-bridge remote-peer statements. One peer statement is needed for each peer, and one peer statement is needed for the local router. In this model, we deployed the use of loopback interfaces to connect the RSRB. Example 13-16 shows the full RSRB configuration for the enterprise router. Example 13-17 shows the RSRB portions of the configuration of the shuttle_15 router.

Example 13-16 RSRB Configuration of the enterprise Router

hostname enterprise ! source-bridge ring-group 100 source-bridge remote-peer 100 tcp 172.16.128.10 Peer for the local router source-bridge remote-peer 100 tcp 172.16.128.5 Peer for the shuttle_5 router source-bridge remote-peer 100 tcp 172.16.128.1 Peer for the shuttle_15 router ! ! interface Loopback20 ip address 172.16.128.10 255.255.255.252 no ip directed-broadcast ! interface Serial0 no ip address no ip directed-broadcast encapsulation frame-relay no ip mroute-cache logging event subif-link-status logging event dlci-status-change frame-relay lmi-type cisco ! interface Serial0.1 multipoint ip address 172.16.2.5 255.255.255.252 no ip directed-broadcast frame-relay map ip 172.16.2.6 170 broadcast ! interface Serial0.2 point-to-point ip address 172.16.2.1 255.255.255.252 no ip directed-broadcast frame-relay interface-dlci 180 ! <<<text omitted>>> ! interface TokenRing0 ip address 172.16.10.1 255.255.255.0 no ip directed-broadcast ring-speed 16 multiring all source-bridge 1 1 100 source-bridge spanning ! router eigrp 2001 network 172.16.0.0 no auto-summary

Example 13-17 RSRB Configuration of the shuttle_15 Router

hostname shuttle_15 ! source-bridge ring-group 100 source-bridge remote-peer 100 tcp 172.16.128.1 Peer for the local router source-bridge remote-peer 100 tcp 172.16.128.5 Peer for the shuttle_5 router source-bridge remote-peer 100 tcp 172.16.128.10 Peer for the enterprise router ! interface Loopback20 ip address 172.16.128.1 255.255.255.252 no ip directed-broadcast ! interface Serial0 ip address 172.16.2.6 255.255.255.252 no ip directed-broadcast encapsulation frame-relay no ip route-cache no ip mroute-cache logging event subif-link-status logging event dlci-status-change frame-relay map ip 172.16.2.5 171 broadcast ! <<<text omitted>>> ! interface TokenRing0 ip address 172.16.1.1 255.255.255.0 no ip directed-broadcast no ip route-cache no ip mroute-cache ring-speed 16 multiring all source-bridge 15 10 100 source-bridge spanning ! router eigrp 2001 network 172.16.0.0 no auto-summary

The RSRB configuration can be verified in the same manner as a normal source-route bridge can. Using the show source-bridge command, verify that the TCP peers are in a closed or open state. The bridge will not transition from closed to open without some type of data sent across it. In this model, the workstation Bones browses the network neighborhood, thereby activating the RSRB. Example 13-18 shows the output of the show source -bridge command performed on the enterprise router.

Example 13-18 show source-bridge Command Output

enterprise# show source-bridge Local Interfaces: receive transmit srn bn trn r p s n max hops cnt cnt drops To0 1 1 100 * f 7 7 7 1019 0 0 Global RSRB Parameters: TCP Queue Length maximum: 100 Ring Group 100: This TCP peer: 172.16.128.10 Maximum output TCP queue length, per peer: 100 Peers: state bg lv pkts_rx pkts_tx expl_gn drops TCP TCP 172.16.128.10 - 3 0 0 0 0 0 TCP 172.16.128.5 open 3 0 1258 1019 0 0 TCP 172.16.128.1 open 3 0 708 1019 346 0 Rings: Rings: bn: 3 rn: 3 remote ma: 4000.30b1.270a TCP 172.16.128.5 fwd: 0 bn: 10 rn: 15 remote ma: 4000.309a.68bb TCP 172.16.128.1 fwd: 0 Explorers: ------- input ------- ------- output ------- spanning all-rings total spanning all-rings total To0 284 735 1019 0 0 0 Explorer fastswitching enabled Local switched: 1019 flushed 0 max Bps 38400 rings inputs bursts throttles output drops To0 1019 0 0 0 enterprise#

Now, modify the previous example to use Frame Relay as the encapsulation type of the RSRB. Figure 13-22 highlights the relevant portions of the network, listing the DLCIs in use.

To configure RSRB to use Frame Relay encapsulation, follow Steps 1 to 3, which are identical to those in the previous section. Frame Relay encapsulation requires only one source-bridge source-bridge remote-peer statement for each remote router connecting the RSRB. In this type of RSRB, you do not configure a remote-peer statement for the local router. Instead, you need to add a Frame-relay map statement for RSRB on the multipoint subinterface. Example 13-19 lists the configuration of the enterprise and shuttle_15 routers, highlighting the Frame Relay RSRB portions. This is the only difference between this model and the TCP model that you just performed.

Example 13-19 RSRB Frame Relay Encapsulation

hostname enterprise ! ip subnet-zero ! source-bridge ring-group 100 source-bridge remote-peer 100 frame-relay interface Serial0.1 170 ! <<<text omitted>>> ! interface Serial0 mtu 4096 no ip address no ip directed-broadcast encapsulation frame-relay no ip mroute-cache logging event subif-link-status logging event dlci-status-change frame-relay lmi-type cisco ! interface Serial0.1 multipoint ip address 172.16.2.5 255.255.255.252 no ip directed-broadcast frame-relay map rsrb 170 broadcast frame-relay map ip 172.16.2.6 170 broadcast ! <<<text omitted>>> ! interface TokenRing0 ip address 172.16.10.1 255.255.255.0 no ip directed-broadcast ring-speed 16 multiring all source-bridge 1 1 100 source-bridge spanning ! _______________________________________________________________________ hostname shuttle_15 ! ip subnet-zero ! source-bridge ring-group 100 source-bridge remote-peer 100 frame-relay interface Serial0 171 ! interface Serial0 mtu 4096 ip address 172.16.2.6 255.255.255.252 no ip directed-broadcast encapsulation frame-relay no ip route-cache no ip mroute-cache logging event subif-link-status logging event dlci-status-change frame-relay map rsrb 171 broadcast frame-relay map ip 172.16.2.5 171 broadcast ! <<<text omitted>>> ! interface TokenRing0 ip address 172.16.1.1 255.255.255.0 no ip directed-broadcast no ip route-cache no ip mroute-cache ring-speed 16 multiring all source-bridge 15 10 100 source-bridge spanning

The status of the RSRB can be viewed with the same show source-bridge command used earlier. Example 13-20 lists the output of this command performed on the enterprise router.

Example 13-20 Status of the RSRB on the enterprise Router

enterprise# show source-bridge Local Interfaces: receive transmit srn bn trn r p s n max hops cnt cnt drops To0 1 1 100 * f 7 7 7 4223 0 0 Global RSRB Parameters: TCP Queue Length maximum: 100 Ring Group 100: No TCP peername set, TCP transport disabled Maximum output TCP queue length, per peer: 100 Peers: state bg lv pkts_rx pkts_tx expl_gn drops TCP FR Serial0.1 170 open 3 0 253 230 16 n/a Rings: bn: 1 rn: 1 local ma: 4007.781a.e789 TokenRing0 fwd: 0 bn: 10 rn: 15 remote ma: 4000.309a.68bb FR Serial0.1 170 fwd: 0 Explorers: ------- input ------- ------- output ------- spanning all-rings total spanning all-rings total To0 1886 2337 4223 0 0 0 Explorer fastswitching enabled Local switched: 4223 flushed 0 max Bps 38400 rings inputs bursts throttles output drops To0 4223 0 0 0 enterprise#

Configuring Other SRB Functions and Features

Cisco provides many useful options for controlling traffic and fine-tuning the source-route bridge environment. Some of the more common features are as follows:

• RSRB TCP and LLC2 local acknowledgment

• Setting largest frame

• Setting Spanning Tree explorers

• Static RIFs

• LSAP and MAC filters

The sections that follow discuss some of these options and how to use them.

RSRB TCP LLC2 Local Acknowledgments

SNA sessions are complete end-to-end sessions. Every frame sent by a front-end processor must be acknowledged by the station or controller receiving the frame. If the SNA session must cross vast geographical distances over low-speed links such as 64 kpbs, there is a high probability of T1 timer expirations. The T1 time is a predefined period of time that a host expects the receiving host to respond, either positively or negatively, to the frame sent to it. All LLC2 frames, including supervisory frames RR, RNR, and REJ, must be acknowledged in an end-to-end manner. Figure 13-23 represents a typical LLC2 session in an RSRB environment.

Cisco offers local acknowledgment for TCP-based RSRBs. Local acknowledgment solves the T1 timer problem without having to change the configuration of the end nodes. With local acknowledgment enabled, all LLC2 frames are acknowledged by the router. The only LLC2 frames that cross the network are I frames, or information frames. Figure 13-24 demonstrates LLC2 local acknowledgment.

To configure local acknowledgment between two RSRBs, use the argument local-ack on the source-bridge remote-peer statement:

Router(config)# source-bridge remote-peer virtual_ring tcp ip_address local-ack

Because the router must maintain a full LLC2 session with every host, the number of simultaneous sessions that it can support could be a factor. Cisco recommends using local acknowledgment only when you are experiencing T1 timer problems or LLC2 problems. Local acknowledgment will not affect NetBIOS timeouts.

Setting the Largest Frame Size

In mixed environments, such as Token Ring and Ethernet, to prevent a lot of segmentation throughout the network, you can fix the largest frame size to 1500 or another value. By setting the frame size to 1500, less segmentation will occur as frames cross Ethernet and Token Ring segments of the network. This easily is accomplished using the lf argument on the Frame Relay, TCP, and FST remote-peer statements:

Router(config)# source-bridge remote-peer virtual_ring frame-relay interface interface_name dlci_number [ lf largest_frame_size ] Router(config)# source-bridge remote-peer virtual_ring tcp ip_address [ lf largest_frame_size ] [ local-ack ] Router(config)# source-bridge remote-peer virtual_ring fst ip_address [ lf largest_frame_size ]

Configuring Spanning Tree Explorers

By default, Cisco routers use all-routes explorer frames to generate the RIF. In large redundant networks, the number of explorers can multiply exponentially as they are duplicated and forwarded throughout the network. Recall from the previous section that Spanning Tree explorers reduce the number of explorers on the network. Spanning Tree nodes will forward Spanning Tree explorers only to nodes that are configured for Spanning Tree. To enable the spanning tree explorers, use the following Token Ring interface command:

Router(config-if)# source-bridge spanning

Microsoft NetBIOS also uses Spanning Tree; therefore, as a rule of thumb in the field, we always configure Spanning Tree explorers when using Microsoft Windows networking.

Configuring a Static RIF

Cisco provides a way to statically configure a RIF on a router. To configure a static RIF, you must be familiar with route control and route descriptor frames. Recall the figures from the previous section of source-route bridging.

Figure 13-25 illustrates the routing control format for the RIF, followed by the descriptions of each field.

• Shaded fields are reserved.

• type ” Explorer type is used, as follows:

  - 00: Specific routes explorer

  - 10: All rings, all-routes explorer

  - 11: All rings, spanning routes (limited broadcast)

• length ” This is the total length in bytes of the RIF.

• D ” This is the direction, indicated as follows:

  - 0: Interpret route left to right (forward)

  - 1: Interpret route right to left (reverse)

• largest ” This is the largest frame that can be handled by this route, as follows:

  - 000: 516 bytes (DDN 1822)

  - 001: 1500 bytes (Ethernet)

  - 010: 2052 bytes

  - 011: 4472 bytes (Token Ring and Cisco maximum)

  - 100: 8144 bytes (Token bus)

  - 101: 11407 bytes

  - 110: 17800 bytes

  - 111: 65535 (initial values)

Figure 13-26 describes the routing descriptor format of the RIF string. When you configure a static RIF, it is in dotted-hexadecimal format.

• Ring number ” Unique decimal ring number within the bridged network.

• Bridge number ” Unique decimal bridge number between any bridges connecting the same two rings. A bridge number of 0 indicates that the RIF is terminating.

Figure 13-27 presents an SRB network.

The static RIF from router_johnson to station Bravo would read as follows:

0830.0072.064a.00b0

Use the previous figures to break down the RIF into its significant components . The 0830 is the 16-bit RC field. Reading from left to right, the bit pattern is as follows:

The first two bits, from left to right again, equal 00. This sets the explorer type to be a specific routes explorer; you want to use this explorer type because this is a static RIF. Bit 3 is set to 0 and is reserved. The next five bits set the length of the RIF in bytes. In this example, the RIF ”not just the ring-bridge-ring part, but the whole RIF ”is eight bytes. The next bit, the D bit, is set to 0, indicating that the RIF is read from left to right, or forward. The next three bits are set to 011, which sets the frame size to be 4472, the Cisco maximum. The last four bits are reserved.

The RD fields, the next three bytes, break down rather easily.

The next three bytes, 0072, 064a, and 00b0, are the three 16-bit RD fields. The first three bits of each byte, are the ring number in hexadecimal format. The last bit is the ring number in hexadecimal format. For the RIF in this example, you have the following:

RING7-BRIDGE2 = 0072 RING100-BRIDGE10 = 064a RING11-BRIDGE0 = 00b0

A bridge of 0 tells the SRB to terminate the RIF and that no more bridges follow the ring.

To configure a static RIF on a router, use the following global router command:

Router(config)# rif mac-address rif-string { interface-name ring-group ring }

Example 13-21 demonstrates the configuration of the static RIF from router_johnson to station bravo.

Example 13-21 Static RIF Configuration

router_johnson(config)# rif 1000.1000.2000 0830.0072.064a.00b0 to0

LSAP, MAC, and NetBIOS Filters

We will discuss LSAP and MAC filters more in the upcoming sections. For now, we simply want to show you the syntax for applying the filters to a source-route bridge environment.

To configure LSAP filters for IEEE 802 encapsulated frames, use the following syntax:

Router(config-if)# source-bridge input-lsap-list access_list_number Router(config-if)# source-bridge output-lsap-list access_list_number Router(config)# rsrb remote-peer ring-group group [ tcp ip_address fst ip_address interface interface_name ] lsap-output-list access_list_number

The LSAP access lists is in the range of 200 to 299 and filters based on LSAP type code.

To filter based on IEEE 802 source addresses, use the following syntax:

Router(config-if)# source-bridge input-address-list access_list_number Router(config-if)# source-bridge output-address-list access_list_number

The access list number ranges from 700 to 799.

To filter based on NetBIOS name, use the following commands:

Router(config-if)# netbios input-access-filter host station_name Router(config-if)# netbios output-access-filter host station_name Router(config)# rsrb remote-peer ring-group [ tcp ip_address fst ip_address interface interface_name ] netbios-output-list access_list_number