Cisco IE 2000 Switch Software Configuration Guide, Cisco IOS Release 15.0(2)EB

FlexLinks

FlexLinks are a pair of a Layer 2 interfaces (switch ports or port channels) where one interface is configured to act as a backup to the other. The feature provides an alternative solution to the Spanning Tree Protocol (STP). Users can disable STP and still retain basic link redundancy. FlexLinks are typically configured in service provider or enterprise networks where customers do not want to run STP on the switch. If the switch is running STP, FlexLinks is not necessary because STP already provides link-level redundancy or backup.

You configure FlexLinks on one Layer 2 interface (the active link) by assigning another Layer 2 interface as the FlexLinks or backup link. When one of the links is up and forwarding traffic, the other link is in standby mode, ready to begin forwarding traffic if the other link shuts down. At any given time, only one of the interfaces is in the linkup state and forwarding traffic. If the primary link shuts down, the standby link starts forwarding traffic. When the active link comes back up, it goes into standby mode and does not forward traffic. STP is disabled on FlexLinks interfaces.

In Figure 24-1, ports 1 and 2 on switch A are connected to uplink switches B and C. Because they are configured as FlexLinks, only one of the interfaces is forwarding traffic; the other is in standby mode. If port 1 is the active link, it begins forwarding traffic between port 1 and switch B; the link between port 2 (the backup link) and switch C is not forwarding traffic. If port 1 goes down, port 2 comes up and starts forwarding traffic to switch C. When port 1 comes back up, it goes into standby mode and does not forward traffic; port 2 continues forwarding traffic.

You can also choose to configure a preemption mechanism, specifying the preferred port for forwarding traffic. For example, in the example in Figure 24-1, you can configure the FlexLinks pair with preemption mode. In the scenario shown, when port 1 comes back up and has more bandwidth than port 2, port 1 begins forwarding traffic after 60 seconds. Port 2 becomes the standby port. You do this by entering the interface configuration switchport backup interface preemption mode bandwidth and switchport backup interface preemption delay commands.

Figure 24-1 FlexLinks Configuration Example

If a primary (forwarding) link goes down, a trap notifies the network management stations. If the standby link goes down, a trap notifies the users.

FlexLinks are supported only on Layer 2 ports and port channels, not on VLANs or on Layer 3 ports.

VLAN FlexLinks Load Balancing and Support

VLAN FlexLinks load-balancing allows you to configure a FlexLinks pair so that both ports simultaneously forward the traffic for some mutually exclusive VLANs. For example, if FlexLinks ports are configured for 1 to100 VLANs, the traffic of the first 50 VLANs can be forwarded on one port and the rest on the other port. If one of the ports fail, the other active port forwards all the traffic. When the failed port comes back up, it resumes forwarding traffic in the preferred VLANs. This way, apart from providing the redundancy, this FlexLinks pair can be used for load balancing. FlexLinks VLAN load balancing does not impose any restrictions on uplink switches.

Figure 24-2 VLAN FlexLinks Load Balancing Configuration Example

FlexLinks Multicast Fast Convergence

FlexLinks Multicast Fast Convergence reduces the multicast traffic convergence time after a FlexLinks failure.

Learning the Other FlexLinks Port as the mrouter Port

In a typical multicast network, there is a querier for each VLAN. A switch deployed at the edge of a network has one of its FlexLinks ports receiving queries. FlexLinks ports are also always forwarding at any given time.

A port that receives queries is added as an mrouter port on the switch. An mrouter port is part of all the multicast groups learned by the switch. After a changeover, queries are received by the other FlexLinks port. The other FlexLinks port is then learned as the mrouter port. After the changeover, multicast traffic flows through the other FlexLinks port. To achieve faster convergence of traffic, both FlexLinks ports are learned as mrouter ports whenever either FlexLinks port is learned as the mrouter port. Both FlexLinks ports are always part of multicast groups.

Though both FlexLinks ports are part of the groups in normal operation mode, all traffic on the backup port is blocked. So the normal multicast data flow is not affected by the addition of the backup port as an mrouter port. When the changeover happens, the backup port is unblocked, allowing the traffic to flow. In this case, the upstream multicast data flows as soon as the backup port is unblocked.

Generating IGMP Reports

When the backup link comes up after the changeover, the upstream new distribution switch does not start forwarding multicast data, because the port on the upstream router, which is connected to the blocked FlexLinks port, is not part of any multicast group. The reports for the multicast groups were not forwarded by the downstream switch because the backup link is blocked. The data does not flow on this port, until it learns the multicast groups, which occurs only after it receives reports.

The reports are sent by hosts when a general query is received, and a general query is sent within 60 seconds in normal scenarios. When the backup link starts forwarding, to achieve faster convergence of multicast data, the downstream switch immediately sends proxy reports for all the learned groups on this port without waiting for a general query.

Leaking IGMP Reports

To achieve multicast traffic convergence with minimal loss, a redundant data path must be set up before the FlexLinks active link goes down. This can be achieved by leaking only IGMP report packets on the FlexLinks backup link. These leaked IGMP report messages are processed by upstream distribution routers, so multicast data traffic gets forwarded to the backup interface. Because all incoming traffic on the backup interface is dropped at the ingress of the access switch, no duplicate multicast traffic is received by the host. When the FlexLinks active link fails, the access switch starts accepting traffic from the backup link immediately. The only disadvantage of this scheme is that it consumes bandwidth on the link between the distribution switches and on the backup link between the distribution and access switches. This feature is disabled by default and can be configured by using the switchport backup interface interface-id multicast fast-convergence command.

When this feature has been enabled at changeover, the switch does not generate the proxy reports on the backup port, which became the forwarding port.

MAC Address-Table Move Update

The MAC address-table move update feature allows the switch to provide rapid bidirectional convergence when a primary (forwarding) link goes down and the standby link begins forwarding traffic.

In Figure 24-3, switch A is an access switch, and ports 1 and 2 on switch A are connected to uplink switches B and D through a FlexLinks pair. Port 1 is forwarding traffic, and port 2 is in the backup state. Traffic from the PC to the server is forwarded from port 1 to port 3. The MAC address of the PC has been learned on port 3 of switch C. Traffic from the server to the PC is forwarded from port 3 to port 1.

If the MAC address-table move update feature is not configured and port 1 goes down, port 2 starts forwarding traffic. However, for a short time, switch C keeps forwarding traffic from the server to the PC through port 3, and the PC does not get the traffic because port 1 is down. If switch C removes the MAC address of the PC on port 3 and relearns it on port 4, traffic can then be forwarded from the server to the PC through port 2.

If the MAC address-table move update feature is configured and enabled on the switches in Figure 24-3 and port 1 goes down, port 2 starts forwarding traffic from the PC to the server. The switch sends a MAC address-table move update packet from port 2. Switch C gets this packet on port 4 and immediately learns the MAC address of the PC on port 4, which reduces the reconvergence time.

You can configure the access switch, switch A, to send MAC address-table move update messages. You can also configure the uplink switches B, C, and D to get and process the MAC address-table move update messages. When switch C gets a MAC address-table move update message from switch A, switch C learns the MAC address of the PC on port 4. Switch C updates the MAC address table, including the forwarding table entry for the PC.

Switch A does not need to wait for the MAC address-table update. The switch detects a failure on port 1 and immediately starts forwarding server traffic from port 2, the new forwarding port. This change occurs in 100 milliseconds (ms). The PC is directly connected to switch A, and the connection status does not change. Switch A does not need to update the PC entry in the MAC address table.

Figure 24-3 MAC Address-Table Move Update Example

Default Settings for FlexLinks and MAC Address-Table Move Update

Configuration Guidelines for FlexLinks and MAC Address-Table Move Update

Follow these guidelines to configure FlexLinks:

• You can configure up to 16 backup links.
• You can configure only one FlexLinks backup link for any active link, and it must be a different interface from the active interface.
• An interface can belong to only one FlexLinks pair. An interface can be a backup link for only one active link. An active link cannot belong to another FlexLinks pair.
• Neither of the links can be a port that belongs to an EtherChannel. However, you can configure two port channels (EtherChannel logical interfaces) as FlexLinks, and you can configure a port channel and a physical interface as FlexLinks, with either the port channel or the physical interface as the active link.
• A backup link does not have to be the same type (Fast Ethernet, Gigabit Ethernet, or port channel) as the active link. However, you should configure both FlexLinks with similar characteristics so that there are no loops or changes in behavior if the standby link begins to forward traffic.
• STP is disabled on FlexLinks ports. A FlexLinks port does not participate in STP, even if the VLANs present on the port are configured for STP. When STP is not enabled, be sure that there are no loops in the configured topology. Once the FlexLinks configurations are removed, STP is reenabled on the ports.

Follow these guidelines to configure VLAN load balancing on the FlexLinks feature:

• For FlexLinks VLAN load balancing, you must choose the preferred VLANs on the backup interface.
• You cannot configure a preemption mechanism and VLAN load balancing for the same FlexLinks pair.

Follow these guidelines to configure the MAC address-table move update feature:

• You can enable and configure this feature on the access switch to send the MAC address-table move updates.
• You can enable and configure this feature on the uplink switches to receive the MAC address-table move updates.

Configuring FlexLinks

Configuring a Preemption Scheme for FlexLinks

Configuring VLAN Load Balancing on FlexLinks

Configuring the MAC Address-Table Move Update Feature

Configuring the MAC Address-Table Move Update Messages

Configuring FlexLinks Port: Examples

These are configuration examples for learning the other FlexLinks port as the mrouter port when FlexLinks is configured, with output for the show interfaces switchport backup command:

This output shows a querier for VLANs 1 and 401, with their queries reaching the switch through the specified port:

Here is output for the show ip igmp snooping mrouter command for VLANs 1 and 401:

Similarly, both FlexLinks ports are part of learned groups. In this example, GigabitEthernet1/1 is a receiver/host in VLAN 1, which is interested in two multicast groups:

When a host responds to the general query, the switch forwards this report on all the mrouter ports. In this example, when a host sends a report for the group 228.1.5.1, it is forwarded only on GigabitEthernet1/1, because the backup port GigabitEthernet1/2 is blocked. When the active link, GigabitEthernet1/1, goes down, the backup port, GigabitEthernet1/2, begins forwarding.

As soon as this port starts forwarding, the switch sends proxy reports for the groups 228.1.5.1 and 228.1.5.2 on behalf of the host. The upstream router learns the groups and starts forwarding multicast data. This is the default behavior of FlexLinks. This behavior changes when the user configures fast convergence using the switchport backup interface GigabitEthernet1/2 multicast fast-convergence command. This example shows how this feature is configured:

This output shows a querier for VLAN 1 and 401 with their queries reaching the switch through the configured port:

This is output for the show ip igmp snooping mrouter command for VLAN 1 and 401:

Similarly, both the FlexLinks ports are a part of the learned groups. In this example, the port is a receiver/host in VLAN 1, which is interested in two multicast groups:

Whenever a host responds to the general query, the switch forwards this report on all the mrouter ports. When you turn on this feature through the command-line port, and when a report is forwarded by the switch on the configured GigabitEthernet1/1, it is also leaked to the backup port GigabitEthernet1/2. The upstream router learns the groups and starts forwarding multicast data, which is dropped at the ingress because the GigabitEthernet1/2 is blocked. When the active link, GigabitEthernet1/1 goes down, the backup port, GigabitEthernet1/2, begins forwarding. You do not need to send any proxy reports because the multicast data is already being forwarded by the upstream router. By leaking reports to the backup port, a redundant multicast path has been set up, and the time taken for the multicast traffic convergence is minimal.

Configuring a Backup Interface: Example

This example shows how to configure an interface with a backup interface and to verify the configuration:

Configuring a Preemption Scheme: Example

This example shows how to configure the preemption mode as forced for a backup interface pair and to verify the configuration:

Configuring VLAN Load Balancing on FlexLinks: Examples

In the following example, VLANs 1 to 50, 60, and 100 to 120 are configured on the switch:

When both interfaces are up, GigabitEthernet1/1 forwards traffic for VLANs 60 and 100 to 120, and GigabitEthernet1/2 forwards traffic for VLANs 1 to 50.

When a FlexLinks interface goes down (LINK_DOWN), VLANs preferred on this interface are moved to the peer interface of the FlexLinks pair. In this example, if interface Gigabit Ethernet1/1 goes down, Gigabit Ethernet1/2 carries all VLANs of the FlexLinks pair.

When a FlexLinks interface comes up, VLANs preferred on this interface are blocked on the peer interface and moved to the forwarding state on the interface that has just come up. In this example, if interface Gigabit Ethernet1/1 comes up, VLANs preferred on this interface are blocked on the peer interface Gigabit Ethernet1/2 and forwarded on Gigabit Ethernet1/1.

Configuring MAC Address-Table Move Update: Example

This example shows how to configure an access switch to send MAC address-table move update messages:

This example shows how to verify the configuration:

Related Documents

Standards

MIBs

RFCs