You can configure an EtherChannel in one of these modes: Port Aggregation Protocol (PAgP), Link Aggregation Control Protocol (LACP), or On. Configure both ends of the EtherChannel in the same mode:

• When you configure one end of an EtherChannel in either PAgP or LACP mode, the system negotiates with the other end of the channel to determine which ports should become active. If the remote port cannot negotiate an EtherChannel, the local port is put into an independent state and continues to carry data traffic as would any other single link. The port configuration does not change, but the port does not participate in the EtherChannel.

• When you configure an EtherChannel in the on mode, no negotiations take place. The switch forces all compatible ports to become active in the EtherChannel. The other end of the channel (on the other switch) must also be configured in the on mode; otherwise, packet loss can occur.

You can create an EtherChannel on a switch, on a single switch in the stack, or on multiple switches in the stack (known as cross-stack EtherChannel).

Figure 2. Single-Switch EtherChannel

Figure 3. Cross-Stack EtherChannel

If a link within an EtherChannel fails, traffic previously carried over that failed link moves to the remaining links within the EtherChannel. If traps are enabled on the switch, a trap is sent for a failure that identifies the switch, the EtherChannel, and the failed link. Inbound broadcast and multicast packets on one link in an EtherChannel are blocked from returning on any other link of the EtherChannel.

Switch ports exchange PAgP packets only with partner ports configured in the auto or desirable modes. Ports configured in the on mode do not exchange PAgP packets.

Both the auto and desirable modes enable ports to negotiate with partner ports to form an EtherChannel based on criteria such as port speed. and for Layer 2 EtherChannels, based on trunk state and VLAN numbers.

Ports can form an EtherChannel when they are in different PAgP modes as long as the modes are compatible. For example:

• A port in the desirable mode can form an EtherChannel with another port that is in the desirable or auto mode.

• A port in the auto mode can form an EtherChannel with another port in the desirable mode.

A port in the auto mode cannot form an EtherChannel with another port that is also in the auto mode because neither port starts PAgP negotiation.

Network devices are classified as PAgP physical learners or aggregate-port learners. A device is a physical learner if it learns addresses by physical ports and directs transmissions based on that knowledge. A device is an aggregate-port learner if it learns addresses by aggregate (logical) ports. The learn method must be configured the same at both ends of the link.

When a device and its partner are both aggregate-port learners, they learn the address on the logical port-channel. The device sends packets to the source by using any of the ports in the EtherChannel. With aggregate-port learning, it is not important on which physical port the packet arrives.

PAgP cannot automatically detect when the partner device is a physical learner and when the local device is an aggregate-port learner. Therefore, you must manually set the learning method on the local device to learn addresses by physical ports. You also must set the load-distribution method to source-based distribution, so that any given source MAC address is always sent on the same physical port.

You also can configure a single port within the group for all transmissions and use other ports for hot-standby. The unused ports in the group can be swapped into operation in just a few seconds if the selected single port loses hardware-signal detection. You can configure which port is always selected for packet transmission by changing its priority with the pagp port-priority interface configuration command. The higher the priority, the more likely that the port will be selected.

Note

The switch supports address learning only on aggregate ports even though the physical-port keyword is provided in the CLI. The pagp learn-method command and the pagp port-priority command have no effect on the switch hardware, but they are required for PAgP interoperability with devices that only support address learning by physical ports, such as the Catalyst 1900 switch. When the link partner of the switch is a physical learner, we recommend that you configure the switch as a physical-port learner by using the pagp learn-method physical-port interface configuration command. Set the load-distribution method based on the source MAC address by using the port-channel load-balance src-mac global configuration command. The switch then sends packets to the physcial learner using the same port in the EtherChannel from which it learned the source address. Only use the pagp learn-method command in this situation.

A virtual switch can be two or more core switches connected by virtual switch links (VSLs) that carry control and data traffic between them. One of the switches is in active mode. The others are in standby mode. For redundancy, remote switches are connected to the virtual switch by remote satellite links (RSLs).

If the VSL between two switches fails, one switch does not know the status of the other. Both switches could change to the active mode, causing a dual-active situation in the network with duplicate configurations (including duplicate IP addresses and bridge identifiers). The network might go down.

To prevent a dual-active situation, the core switches send PAgP protocol data units (PDUs) through the RSLs to the remote switches. The PAgP PDUs identify the active switch, and the remote switches forward the PDUs to core switches so that the core switches are in sync. If the active switch fails or resets, the standby switch takes over as the active switch. If the VSL goes down, one core switch knows the status of the other and does not change its state.

The Dynamic Trunking Protocol (DTP) and the Cisco Discovery Protocol (CDP) send and receive packets over the physical ports in the EtherChannel. Trunk ports send and receive PAgP protocol data units (PDUs) on the lowest numbered VLAN.

In Layer 2 EtherChannels, the first port in the channel that comes up provides its MAC address to the EtherChannel. If this port is removed from the bundle, one of the remaining ports in the bundle provides its MAC address to the EtherChannel.

PAgP sends and receives PAgP PDUs only from ports that are up and have PAgP enabled for the auto or desirable mode.

Both the active and passive LACP modes enable ports to negotiate with partner ports to an EtherChannel based on criteria such as port speed, and for Layer 2 EtherChannels, based on trunk state and VLAN numbers.

Ports can form an EtherChannel when they are in different LACP modes as long as the modes are compatible. For example:

• A port in the active mode can form an EtherChannel with another port that is in the active or passive mode.

• A port in the passive mode cannot form an EtherChannel with another port that is also in the passive mode because neither port starts LACP negotiation.

The DTP and the CDP send and receive packets over the physical ports in the EtherChannel. Trunk ports send and receive LACP PDUs on the lowest numbered VLAN.

In Layer 2 EtherChannels, the first port in the channel that comes up provides its MAC address to the EtherChannel. If this port is removed from the bundle, one of the remaining ports in the bundle provides its MAC address to the EtherChannel.

LACP sends and receives LACP PDUs only from ports that are up and have LACP enabled for the active or passive mode.

With source-MAC address forwarding, when packets are forwarded to an EtherChannel, they are distributed across the ports in the channel based on the source-MAC address of the incoming packet. Therefore, to provide load-balancing, packets from different hosts use different ports in the channel, but packets from the same host use the same port in the channel.

With destination-MAC address forwarding, when packets are forwarded to an EtherChannel, they are distributed across the ports in the channel based on the destination host’s MAC address of the incoming packet. Therefore, packets to the same destination are forwarded over the same port, and packets to a different destination are sent on a different port in the channel.

With source-and-destination MAC address forwarding, when packets are forwarded to an EtherChannel, they are distributed across the ports in the channel based on both the source and destination MAC addresses. This forwarding method, a combination source-MAC and destination-MAC address forwarding methods of load distribution, can be used if it is not clear whether source-MAC or destination-MAC address forwarding is better suited on a particular switch. With source-and-destination MAC-address forwarding, packets sent from host A to host B, host A to host C, and host C to host B could all use different ports in the channel.

With source-IP address-based forwarding, packets are distributed across the ports in the EtherChannel based on the source-IP address of the incoming packet. To provide load balancing, packets from different IP addresses use different ports in the channel, and packets from the same IP address use the same port in the channel.

With destination-IP address-based forwarding, packets are distributed across the ports in the EtherChannel based on the destination-IP address of the incoming packet. To provide load balancing, packets from the same IP source address sent to different IP destination addresses could be sent on different ports in the channel. Packets sent from different source IP addresses to the same destination IP address are always sent on the same port in the channel.

With source-and-destination IP address-based forwarding, packets are distributed across the ports in the EtherChannel based on both the source and destination IP addresses of the incoming packet. This forwarding method, a combination of source-IP and destination-IP address-based forwarding, can be used if it is not clear whether source-IP or destination-IP address-based forwarding is better suited on a particular switch. In this method, packets sent from the IP address A to IP address B, from IP address A to IP address C, and from IP address C to IP address B could all use different ports in the channel.

Different load-balancing methods have different advantages, and the choice of a particular load-balancing method should be based on the position of the switch in the network and the kind of traffic that needs to be load-distributed.

Figure 5. Load Distribution and Forwarding Methods. In the following figure, an EtherChannel of four workstations communicates with a router. Because the router is a single MAC-address device, source-based forwarding on the switch EtherChannel ensures that the switch uses all available bandwidth to the router. The router is configured for destination-based forwarding because the large number of workstations ensures that the traffic is evenly distributed from the router EtherChannel.

Use the option that provides the greatest variety in your configuration. For example, if the traffic on a channel is going only to a single MAC address, using the destination-MAC address always chooses the same link in the channel. Using source addresses or IP addresses might result in better load-balancing.

With PAgP, if the active switch fails or leaves the stack, the standby switch becomes the new active switch. The new active switch synchronizes the configuration of the stack members to that of the active switch. The PAgP configuration is not affected after an active switch change unless the EtherChannel has ports residing on the old active switch.

With LACP, the system ID uses the stack MAC address from the active switch, and if the active switch changes, the LACP system ID can change. If the LACP system ID changes, the entire EtherChannel will flap, and there will be an STP reconvergence. Use the stack-mac persistent timer command to control whether or not the stack MAC address changes during a active switch failover.

When configuring Layer 2 EtherChannels, follow these guidelines:

• Assign all ports in the EtherChannel to the same VLAN, or configure them as trunks. Ports with different native VLANs cannot form an EtherChannel.

• An EtherChannel supports the same allowed range of VLANs on all the ports in a trunking Layer 2 EtherChannel. If the allowed range of VLANs is not the same, the ports do not form an EtherChannel even when PAgP is set to the auto or desirable mode.

• Ports with different spanning-tree path costs can form an EtherChannel if they are otherwise compatibly configured. Setting different spanning-tree path costs does not, by itself, make ports incompatible for the formation of an EtherChannel.

• For Layer 3 EtherChannels, assign the Layer 3 address to the port-channel logical interface, not to the physical ports in the channel.

When configuring Layer 3 EtherChannels, you should first manually create the port-channel logical interface by using the interface port-channel global configuration command. Then put the logical interface into the channel group by using the channel-group interface configuration command.

Note	To move an IP address from a physical port to an EtherChannel, you must delete the IP address from the physical port before configuring it on the port-channel interface.

Follow these steps to create a port-channel interface for a Layer 3 EtherChannel. This procedure is required.

1.    enable


2.    configure terminal


3.    interface port-channel port-channel-number


4.    no switchport


5.    ip address ip-address mask


6.    end

Switch> enable

Switch# configure terminal

Switch(config)# interface port-channel 5

Switch(config-if)# no switchport

 

Switch(config-if)# ip address ip address 172.10.20.10 255.255.255.0

Switch(config-if)# end

This example shows how to create the logical port channel 5 and assign 172.10.20.10 as its IP address:

Switch# configure terminal
Switch(config)# interface port-channel 5
Switch(config-if)# no switchport
Switch(config-if)# ip address 172.10.20.10 255.255.255.0
Switch(config-if)# end

1.    enable


2.    configure terminal


3.    interface interface-id


4.    no ip address


5.    no switchport


6.    channel-group channel-group-number mode { auto [ non-silent ] | desirable [ non-silent ] | on } | { active | passive }


7.    end

Switch> enable

Switch# configure terminal

Switch(config)# interface gigabitethernet 1/0/2

Switch(config-if)# no ip address

Switch(config-if)# no switchport

Switch(config-if)# channel-group 5 mode auto
               

Switch(config-if)# end

1.    enable


2.    configure terminal


3.    lacp system-priority priority


4.    end

Switch> enable

Switch# configure terminal

Switch(config)# lacp system-priority 32000

Switch(config)# end

By default, all ports use the same port priority. If the local system has a lower value for the system priority and the system ID than the remote system, you can affect which of the hot-standby links become active first by changing the port priority of LACP EtherChannel ports to a lower value than the default. The hot-standby ports that have lower port numbers become active in the channel first. You can use the show etherchannel summary privileged EXEC command to see which ports are in the hot-standby mode (denoted with an H port-state flag).

Note

If LACP is not able to aggregate all the ports that are compatible (for example, the remote system might have more restrictive hardware limitations), all the ports that cannot be actively included in the EtherChannel are put in the hot-standby state and are used only if one of the channeled ports fails.

Follow these steps to configure the LACP port priority. This procedure is optional.

1.    enable


2.    configure terminal


3.    interface interface-id


4.    lacp port-priority priority


5.    end

Switch> enable

Switch# configure terminal

 

Switch(config)# interface gigabitethernet 1/0/2

Switch(config-if)# lacp port-priority 32000

Switch(config-if)# end