NVIDIA Onyx User Manual v3.10.4006
NVIDIA Onyx User Manual v3.10.4006

Store-and-Forward

Store-and-Forward is used to describe a functionality where a switch receives a complete packet, stores it, and only then forwards it.

since the switch make forwarding decisions based on the destination address which is at the header of the packet, the switch can make the forwarding decision before receiving the complete packet, this process is called cut-through, the switch forwards part of the packet before receiving the complete packet.

Cut-through allows lower latency and saves buffer space, but if an error occurred in the packet while utilizing cut-through, the packet will be forwarded with an error, alternatively, utilizing store-and-forward allows the switch to drop erroneous packets.

The standard implementation of forwarding mode is for the entire switch; either all ports on a switch are in store-and-forward mode or all ports on a switch are in cut-through mode. NVIDIA implements forwarding mode per egress port, which is a more flexible method and vital in cases where a switch is connected to both a storage device and a compute server among other setups.

For more information about this feature and its potential applications, please refer to the following community post:

switchmode store-and-forward

switchmode store-and-forward

no switchmode store-and-forward

disable switchmode store-and-forward

Enables global store-and-forward configuration on the switch.

The no form of the command removes store-and-forward configuration from the switch and reverts it back to the switch’s global configuration.

The disable form of the command configures the forwarding mode to cut-through.

Syntax Description

N/A

Default

N/A

Configuration Mode

config

config interface ethernet

config interface port-channel

config interface mlag-port-channel

History

3.6.3640

3.6.6000

Added “config interface mlag-port-channel” configuration mode

Example

switch (config)# switchmode store-and-forward

Related Commands

Notes


© Copyright 2023, NVIDIA. Last updated on Mar 5, 2024.