DCB Willing vs. Non-willing Modes

The concept of willing vs. non-willing is in regards to the DCBx negotiation to arrive at a set of DCB settings. It refers to which link-partner will be the lead (non-willing) and the follower (willing). The lead is the one that proposes DCB configuration via LLDP frames and the follower is the one that is "willing" to apply it to its own configuration.

• The willing / non-willing designation only matters if there is a DCBx negotiation occuring.
• DCB settings can be manually set without DCBx being used.

A common strategy for using willing and non-willing modes in a cluster:

1. Set switches as non-willing.
2. Configure DCB (priority settings, traffic classes, bandwidth allocations, etc.) on the switch ports.
3. Set adapters as willing.
4. Adapters are automatically configured.

This helps simplify DCB cluster configuration by centralizing DCB settings on a switch and pushing the configuration to the adapters (rather than configuring each host individually).

Priority flow control (PFC) is supported in both willing and non-willing modes on Intel® Ethernet 800 Series. This series also has two DCB modes: software and firmware. For more background on software and firmware modes, refer to the ice driver README.

• For PFC willing mode, software DCB is recommended but firmware DCB is also supported.
• For PFC non-willing mode, software DCB must be used.