Eth-Trunk

What Is Eth-Trunk?

Eth-Trunk aggregates numerous physical links into a logical link, boosting link bandwidth. The dynamic backup among the aggregated links significantly enhances overall link reliability.

Why Do We Need Eth-Trunk?

As network infrastructures expand, the demand for backbone links with enhanced bandwidth and reliability becomes crucial. Traditional approaches involve using high-speed cards or compatible devices, incurring significant costs and lacking flexibility.

Eth-Trunk, or Ethernet link aggregation, addresses this by bundling multiple physical links into a logical link, effectively increasing link bandwidth. This technology goes further by substantially enhancing link reliability through a backup mechanism. Each Link Aggregation Group (LAG) corresponds to a distinct logical interface, referred to as the link aggregation interface or Eth-Trunk interface. This interface, akin to a physical Ethernet interface, distinguishes itself by the need to select one or more member interfaces for traffic forwarding.

The Eth-Trunk resides within the data link layer, positioned between the MAC sub-layer and the LLC sub-layer. Within this layer, the Eth-Trunk module manages a forwarding table with primary fields:

• Hash key: Computed through a hash algorithm based on the MAC or IP address within a data packet.

• Interface number: The Eth-Trunk forwarding entries are constrained by the maximum member interfaces in an Eth-Trunk. Distinct hash keys correspond to different outbound interfaces.

The Eth-Trunk module utilizes this forwarding table to execute the following data frame forwarding process:

1. The Eth-Trunk module receives a data frame from the MAC sub-layer, extracting either its source MAC/IP address or destination MAC/IP address.

2. Subsequently, the Eth-Trunk module calculates the hash key using the designated hash algorithm.

3. With the computed hash key, the Eth-Trunk module searches the Eth-Trunk forwarding table for the associated interface, facilitating the transmission of the data frame through that interface.

What Modes Does Eth-Trunk Support?

Eth-Trunk functionality is adaptable and can operate in either manual or Link Aggregation Control Protocol (LACP) mode, contingent upon the activation of LACP.

- Eth-Trunk in Manual Mode:

In manual mode, the creation of an Eth-Trunk and the addition of member interfaces are manual processes, devoid of LACP requirements. Load balancing is achieved among active links, and if one link becomes inactive, traffic seamlessly redistributes across the remaining active links. This mode is particularly suitable for scenarios where a higher bandwidth between two directly-connected devices is needed, and LACP support is absent or limited.

For instance, envision an Eth-Trunk established between DeviceA and DeviceB in manual mode. All three active links collaboratively balance and transmit traffic. If one link encounters a fault, the remaining two links seamlessly share the traffic load.

Eth-Trunk in manual mode

- Eth-Trunk in LACP Mode:

In LACP mode, compliant with IEEE 802.3ad standards, dynamic link aggregation and de-aggregation are facilitated. LACP-enabled devices at both ends engage in the creation of an Eth-Trunk, exchanging Link Aggregation Control Protocol Data Units (LACPDUs). Subsequently, LACP assumes responsibility for Eth-Trunk maintenance and adapts member links when there are alterations in link aggregation conditions.

Consider a scenario where four interfaces on DeviceA are intended for bundling into an Eth-Trunk and connecting to corresponding interfaces on DeviceB. However, a potential fault arises as one interface on DeviceA is erroneously linked to an interface on DeviceC. In manual mode, such a fault could go unnoticed.

In contrast, if LACP is enabled on both DeviceA and DeviceB, the Eth-Trunk selectively designates active links (those linked to DeviceB) to transmit data following negotiation. Consequently, data sent by DeviceA destined for DeviceB exclusively reaches its intended destination.

Incorrect Eth-Trunk connection

The following table compares the manual mode and LACP mode.

Application Scenarios of Eth-Trunk

The core switch DeviceA in the data center connects to access switches DeviceB and DeviceC, with the latter serving numerous users. Both DeviceB and DeviceC link to multiple users. External connectivity to the network beyond the data center is facilitated by DeviceA through the egress router.

As user communication intensifies due to a growing user base, it becomes imperative to enhance the links between DeviceA and DeviceB, as well as DeviceA and DeviceC. These links must not only offer ample bandwidth to accommodate increased user traffic but also guarantee a high level of reliability. To meet these requirements, Eth-Trunk1 is established between DeviceA and DeviceB, while Eth-Trunk2 is deployed between DeviceA and DeviceC.

Networking Typology of the Eth-Trunk

The operational mode for Eth-Trunk can be selected based on the following scenarios:

• Opt for LACP mode if both devices connected to the Eth-Trunk support LACP.

• In cases where LACP is not supported by either device connected to the Eth-Trunk, the manual mode becomes the necessary choice.