No relevant resource is found in the selected language.

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>Search


To have a better experience, please upgrade your IE browser.


CX11x, CX31x, CX710 (Earlier Than V6.03), and CX91x Series Switch Modules V100R001C10 Configuration Guide 13

The documents describe the configuration of various services supported by the CX11x&CX31x&CX91x series switch modules The description covers configuration examples and function configurations.
Rate and give feedback:
Huawei uses machine translation combined with human proofreading to translate this document to different languages in order to help you better understand the content of this document. Note: Even the most advanced machine translation cannot match the quality of professional translators. Huawei shall not bear any responsibility for translation accuracy and it is recommended that you refer to the English document (a link for which has been provided).



As shown in Figure 5-18, DeviceA and DeviceB are connected through three Ethernet physical links. These three Ethernet physical links are bundled into an Eth-Trunk link. The bandwidth of the Eth-Trunk link is the sum of bandwidth of the three Ethernet physical links, so bandwidth is increased. The three Ethernet physical links back up each other, which improves reliability.

The rates at both ends of the Eth-Trunk must be the same. It is recommended that the number of connected physical interfaces, and jumbo and flow control configuration at both ends of the Eth-Trunk be the same.
Figure 5-18 Eth-Trunk networking

The link aggregation interface can be used as a common Ethernet interface to implement routing protocols and other services. Unlike a common Ethernet interface, the link aggregation interface uses one or more member interfaces to forward data.

Link aggregation concepts are described as follows:

  • Link aggregation, LAG (Link Aggregation Group), and link aggregation interface

    Link aggregation is a method of bundling a group of physical interfaces into a logical interface to increase bandwidth and improve reliability.

    An LAG is the logical link bundled by many Ethernet links.

    Each LAG corresponds to a logical interface, that is, link aggregation interface or Eth-Trunk.

  • Member interface and member link

    The interfaces that constitute an Eth-Trunk are member interfaces. The link corresponding to a member interface is member link.

  • Active and inactive interfaces and links

    Member interfaces can be classified into active interfaces, which forward data and inactive interfaces, which do not.

    Links connected to active interfaces are called active links, and links connected to inactive interfaces are called inactive links.

  • Upper threshold for the number of active interfaces

    When the number of active interfaces reaches this threshold, the bandwidth of the Eth-Trunk will not increase even if more member links go Up. This guarantees higher network reliability. When the number of active member interfaces reaches the upper threshold, additional active member interfaces are set to Down.

    For example, 8 trouble-free member links are bundled into a trunk link, each with a bandwidth of 1 Gbit/s. The trunk link, however, only needs to provide a maximum bandwidth of 5 Gbit/s. By setting the maximum number of Up member links to 5 or a greater, any unselected Up links automatically enter backup status, improving reliability.

    The upper threshold for the number of active interfaces is inapplicable to the manual load balancing mode. For details about the manual load balancing mode, see Link Aggregation in Manual Load Balancing Mode.

  • Lower threshold for the number of active interfaces

    When the number of active interfaces falls below this threshold, the trunk interface goes Down. This guarantees the trunk interface a minimum available bandwidth.

    For example, if the trunk interface is required to provide a minimum bandwidth of 2 Gbit/s and each member link's bandwidth is 1 Gbit/s, the minimum number of Up member links must be set to 2 or a greater.

Forwarding Principle

As shown in Figure 5-19, an Eth-Trunk link is deployed on the data link layer, that is, between the physical layer and the MAC sub-layer.

Figure 5-19 Eth-Trunk interface in the Ethernet protocol stack

An Eth-Trunk interface is assumed to be a physical interface on the MAC sub-layer. Therefore, frames transmitted in the MAC sub-layer only need to be delivered to the Eth-Trunk module that maintains an Eth-Trunk forwarding table.

The Eth-Trunk forwarding table is composed of the following parts:

  • HASH-KEY value

    The key value is calculated through the hash algorithm on the MAC address or IP address in the packet.

  • Interface number

    Eth-Trunk forwarding entries are relevant to the number of member interfaces in the Eth-Trunk. Different HASH-KEY values are mapped to different outbound interfaces.

    For example, an Eth-Trunk supports a maximum of 16 member interfaces. If four physical interfaces, 1, 2, 3, and 4, are bundled into an Eth-Trunk, the Eth-Trunk forwarding table contains four entries, and the HASH-KEY values correspond to interface numbers as shown in Figure 5-20.

Figure 5-20 Example of an Eth-Trunk forwarding table

The Eth-Trunk module forwards a frame according to the Eth-Trunk forwarding table. The forwarding process is as follows:

  1. The Eth-Trunk module receives a frame from the MAC sub-layer, and then extracts its source MAC address/IP address or destination MAC address/IP address.

  2. The Eth-Trunk module calculates the HASH-KEY value using the hash algorithm.

  3. Based on the HASH-KEY value, the Eth-Trunk module searches the Eth-Trunk forwarding table for the interface number, and then sends the frame from the corresponding interface.

Link Aggregation in Manual Load Balancing Mode

Link aggregation can work in manual load balancing mode and LACP mode.

In manual load balancing mode, you need to manually create an Eth-Trunk interface and add member interfaces to the Eth-Trunk interface, without the assistance of the LACP protocol. In this mode, all active links load balance the traffic evenly. If an active link fails, the other active links share the traffic evenly. If a high link bandwidth between two directly connected devices is required but the peer device does not support the LACP protocol, you can use the manual load balancing mode.

Link Aggregation in LACP Mode


Eth-Trunk in manual load balancing mode, as a link aggregation technology, can increase the bandwidth. However, this mode can only detect link disconnections, but cannot detect other faults such as link layer faults and incorrect link connections.

The Link Aggregation Control Protocol (LACP) is used, which can improve fault tolerance of the Eth-Trunk and ensure high reliability of the member links.

LACP uses a standard negotiation mechanism for switching devices, ensuring that switching devices automatically create and enable aggregated links based on their configurations. After aggregated links are created, LACP maintains link status. If an aggregated link's status changes, LACP automatically adjusts or disables the link.

For example, in Figure 5-21 an Eth-Trunk link should be established between DeviceA and DeviceB by bundling four interfaces on DeviceA into an Eth-Trunk interface and connecting it to the corresponding interfaces on DeviceB. However, one of the interfaces is incorrectly connected to the interface on DeviceC. As a result, the Eth-Trunk in manual load balancing mode cannot detect the fault in time and continues sending data to DeviceC.

If LACP is enabled on DeviceA and DeviceB, the Eth-Trunk correctly selects active links to forward data after negotiation. Data sent by DeviceA can reach DeviceB.

Figure 5-21 Incorrect Eth-Trunk connection


  • LACP system priority

    In LACP mode, active member interfaces selected by both devices must be consistent; otherwise, the LAG cannot be established. To keep active member interfaces consistent at both ends, set a higher priority for one end. In this manner, the other end selects active member interfaces based on the selection of the peer. LACP system priorities are set on devices at both ends of a trunk link. The smaller the LACP system priority value, the higher the LACP system priority.

  • LACP interface priority

    Interface LACP priorities are set to prioritize interfaces of an Eth-Trunk. Interfaces with higher priorities are selected as active interfaces. The smaller the LACP interface priority value, the higher the LACP interface priority.

Implementation of Link Aggregation in LACP Mode

LACP, as specified in IEEE 802.3ad, implements dynamic link aggregation and de-aggregation, allowing both ends to exchange LACPDUs.

After member interfaces are added to the Eth-Trunk interface in LACP mode, each end sends LACPDUs to inform its peer of its system priority, MAC address, member interface priorities, interface numbers, and keys. Keys are used to determine whether the remote end connected to each interface is in the same LAG and whether bandwidth of each interface is the same. After being informed, the peer compares this information with that saved on itself, and selects which interfaces to be aggregated. Then, LACP negotiation occurs, selecting the active interfaces and links.

  • The process of setting up an Eth-Trunk link in LACP mode is as follows:

    1. Devices at both ends send LACPDUs to each other.

      As shown in Figure 5-22, you need to manually create an Eth-Trunk link in LACP mode on DeviceA and DeviceB and add member interfaces to the Eth-Trunk. Then the member interfaces are enabled with LACP, and devices at both ends can send LACPDUs to each other.

      Figure 5-22 LACPDUs sent in LACP mode network diagram
    2. Determine the Actor and active links.

      As shown in Figure 5-23, devices at both ends receive LACPDUs from each other. For example, when DeviceB receives LACPDUs from DeviceA, DeviceB checks and records information about DeviceA and compares system priorities. If the system priority of DeviceA is higher than that of DeviceB, DeviceA acts as the Actor.

      After devices at both ends select the Actor, they select active interfaces according to the priorities of the Actor's interfaces. Then active interfaces are selected, active links in the LAG are specified, and data is forwarded by active links.

      Figure 5-23 Actor selection process in LACP mode network diagram
  • LACP preemption

    After LACP preemption is enabled, interfaces with higher priorities in an LAG function as active interfaces.

    As shown in Figure 5-24, port1, port2, and port3 are member interfaces of an Eth-Trunk. DeviceA acts as the Actor. The upper threshold for the number of active interfaces is 2. LACP priorities of port1, port2, and port3 are 10, 20, and 30 respectively. When LACP negotiation is complete, port1 and port2 are selected as active interfaces because their LACP priorities are higher, and port3 is selected as the backup interface.

    Figure 5-24 LACP preemption

    LACP preemption is typically enabled in the following situations:

    • port1 becomes faulty, and then recovers. When port1 fails, port3 replaces port1 to transmit services. After port1 recovers, by default, port1 still retains in backup state. If LACP preemption is enabled on Eth-Trunk, port1 becomes the active interface and port3 becomes the backup interface.
    • If LACP preemption is enabled and port3 needs to replace port1 or port2 to become the active interface, set the highest LACP priority value for port3. If LACP preemption is not enabled, the system neither re-selects the active interface nor switches the active interface when the priority of a backup interface is higher than that of the active interface.
  • LACP preemption delay

    After LACP preemption occurs, the backup link waits for a set period of time before switching to active status. This period is called LACP preemption delay. The LACP preemption delay is set to prevent unstable data transmission along Eth-Trunk links caused by frequent status changes in member links.

  • Switchover between active links and inactive links

    In LACP mode, a link switchover in the LAG is triggered if a device at one end detects one of the following events:

    • An active link goes Down.

    • LACP detects a link fault.

    • An active interface becomes unavailable.

    • If LACP preemption is enabled, the backup interface's priority is changed to be higher than that of the current active interface.

    When any of the preceding triggering conditions is met, the link switchover is performed in the following steps:

    1. The faulty link is disabled.

    2. The highest priority backup link is selected to replace the faulty active link.

    3. The highest priority backup link becomes the active link and begins forwarding data.

LACP Implementation Modes

LACP can work in static or dynamic LACP mode:

  • Static LACP mode

    In static LACP mode, two ends exchange LACP packets to negotiate link aggregation parameters to determine active and inactive interfaces.

    In static LACP mode, you must manually create an Eth-Trunk interface and add member interfaces to the Eth-Trunk interface. Different from link aggregation in manual load balancing mode, active member interfaces are selected by sending LACP data units (LACPDUs) in static LACP mode. That is, when a group of interfaces are added to an Eth-Trunk interface, devices at the two ends determine active interfaces and inactive interfaces by sending LACPDUs to each other.

    The static LACP mode is called the M:N mode. In this mode, both load balancing and redundancy backup can be implemented. In the link aggregation group, M pieces of links are active to forward data and perform load balancing and other N pieces of links are inactive. The inactive links function as standby links and do not forward data. When one active link fails, the system selects the link with the highest priority from the standby links to replace the faulty link. Then, the link becomes active and starts to forward data.

    On the network shown in Figure 5-25, DeviceA and DeviceB are directly connected. Both Devices support LACP. Eth-Trunk interfaces working in static LACP mode can be configured on the two Devices to implement load balancing and link backup. Static LACP mode is mainly applied in situations where the bandwidth of M links must be assured, and a fault tolerance mechanism in place. If an active link fails, the system selects the backup link with the highest priority and this backup link becomes the active link.

    Figure 5-25 Schematic diagram for Eth-Trunk interfaces in static LACP mode

  • Dynamic LACP mode

    LACPDU exchange in static and dynamic LACP modes is the same, but processing upon LACP negotiation failure is different:
    • In static LACP mode, the Eth-Trunk becomes Down and cannot forward data after LACP negotiation failure.
    • In dynamic LACP mode, the Eth-Trunk becomes Down after LACP negotiation failure, but its member interfaces inherit VLAN attributes of the Eth-Trunk and enter the Indep state. The member interfaces can forward data at Layer 2.

    After a device deployed with an Eth-Trunk interface in dynamic LACP mode receives LACPDUs from a remote end, the two devices will use LACPDUs to negotiate link aggregation parameters. After the negotiation, the link aggregation function will be the same as that of Eth-Trunk interfaces working in static LACP mode.

    An Eth-Trunk in dynamic LACP mode is often used to directly connect the device and server. As shown in Figure 5-26, ServerA needs to obtain the configuration file from ServerB through DeviceA.
    • After ServerA restarts and has no configuration, LACP negotiation fails. The dynamic LACP mode can ensure that ServerA obtains the configuration file from ServerB through the Eth-Trunk member interface.
    • After DeviceA receives LACPDUs from ServerA, it uses LACPDUs to negotiate link aggregation parameters with ServerA.
    Figure 5-26 Schematic diagram for the Eth-Trunk interface in dynamic LACP mode

Load Balancing Using Link Aggregation

A data flow is a group of data packets with one or more identical attributes. The attributes refer to the source MAC address, destination MAC address, source IP address, destination IP address, source TCP/UDP port number, and destination TCP/UDP port number.

Because there are multiple physical links between devices of the Eth-Trunk, the first data frame of the same data flow is transmitted on one physical link, and the second data frame may be transmitted on another physical link. In this case, the second data frame may arrive at the peer device earlier than the first data frame. As a result, packet mis-sequencing occurs.

To prevent packet mis-sequencing, Eth-Trunk uses the load balancing mechanism. This mechanism uses the hash algorithm to calculate the address in a data frame and generates a hash key value. Then the system searches for the outbound interface in the Eth-Trunk forwarding table based on the generated hash key value. Each MAC or IP address corresponds to a hash key value, so the system uses different outbound interfaces to forward data. This mechanism ensures that frames of the same data flow are forwarded on the same physical link and implements flow-based load balancing. Flow-based load balancing ensures the sequence of data transmission, but reduces the bandwidth usage.

Preferentially Forwarding Local Interface Traffic on an Eth-Trunk of a stack Device


  • Stack device

    The stack device is a logical device formed by connecting multiple devices through stack cables. In Figure 5-27, DeviceB and DeviceC are connected to form a logical device.

  • Inter-device Eth-Trunk

    Physical interfaces in the stack are added to an Eth-Trunk. When a device in the stack fails or a device physical interface added to the Eth-Trunk fails, traffic can be transmitted between devices through stack cables. This ensures reliable transmission and implements device backup.

  • Preferential forwarding of local interface traffic

    In b of Figure 5-27, traffic from DeviceB or DeviceC is only forwarded through local member interfaces when the network runs properly. In a of Figure 5-27, traffic is forwarded across devices through stack cables.

    Figure 5-27 Inter-device Eth-Trunk

Inter-Device Eth-Trunk Supporting Preferential Forwarding of Local Interface Traffic

In a stack, an Eth-Trunk is configured to be the outbound interface of traffic to ensure reliable transmission. Member interfaces of the Eth-Trunk are located on different devices. When the stack device forwards traffic, the Eth-Trunk may select an inter-device member interface based on the hash algorithm. This occupies bandwidth resources between devices and reduces traffic forwarding efficiency.

As shown in Figure 5-27, DeviceB and DeviceC constitute a stack, and the stack connects to DeviceA through an Eth-Trunk. After the Eth-Trunk in the stack is configured to preferentially forward local interface traffic, the following functions are implemented:

  • Forwarding received traffic by the local device

    When DeviceB has member interfaces of the Eth-Trunk and the member interfaces function properly, the Eth-Trunk forwarding table of DeviceB contains only local member interfaces. In this manner, the hash algorithm selects a local member interface, and traffic is only forwarded through DeviceB.

  • Forwarding received traffic by another device

    When DeviceB does not have any member interface of the Eth-Trunk or all member interfaces are faulty, the Eth-Trunk forwarding table of DeviceB contains all available member interfaces. In this manner, the hash algorithm selects a member interface on DeviceC, and traffic is forwarded through DeviceC.

  • This function is only valid for known unicast packets, and is invalid for unknown unicast packets, broadcast packets and multicast packets.
  • Before configuring an Eth-Trunk to preferential forward local interface traffic, ensure that member interfaces of the local Eth-Trunk have sufficient bandwidth to forward local traffic; otherwise, traffic may be discarded.
Updated: 2019-12-13

Document ID: EDOC1000041694

Views: 61087

Downloads: 3623

Average rating:
This Document Applies to these Products
Related Version
Related Documents
Previous Next