Service Manual

Conguring Priority-Based Flow Control
Priority-Based Flow Control (PFC) provides a ow control mechanism based on the 802.1p priorities in converged Ethernet trac
received on an interface and is enabled by default when you enable DCB.
As an enhancement to the existing Ethernet pause mechanism, PFC stops trac transmission for specied priorities (Class of
Service (CoS) values) without impacting other priority classes. Dierent trac types are assigned to dierent priority classes.
When trac congestion occurs, PFC sends a pause frame to a peer device with the CoS priority values of the trac that is to be
stopped. Data Center Bridging Exchange protocol (DCBx) provides the link-level exchange of PFC parameters between peer
devices. PFC allows network administrators to create zero-loss links for Storage Area Network (SAN) trac that requires no-drop
service, while retaining packet-drop congestion management for Local Area Network (LAN) trac.
To congure PFC, follow these steps:
1. Create a DCB Map.
CONFIGURATION mode
dcb-map dcb-map-name
The dcb-map-name variable can have a maximum of 32 characters.
2. Create a PFC group.
CONFIGURATION mode
priority-group group-num {bandwidth bandwidth | strict-priority} pfc on
The range for priority group is from 0 to 7.
Set the bandwidth in percentage. The percentage range is from 1 to 100% in units of 1%.
Committed and peak bandwidth is in megabits per second. The range is from 0 to 40000.
Committed and peak burst size is in kilobytes. Default is 50. The range is from 0 to 10000.
The pfc on command enables priority-based ow control.
3. Specify the dot1p priority-to-priority group mapping for each priority.
priority-pgid dot1p0_group_num dot1p1_group_num ...dot1p7_group_num
Priority group range is from 0 to 7. All priorities that map to the same queue must be in the same priority group.
Leave a space between each priority group number. For example: priority-pgid 0 0 0 1 2 4 4 4 in which priority group 0 maps to
dot1p priorities 0, 1, and 2; priority group 1 maps to dot1p priority 3; priority group 2 maps to dot1p priority 4; priority group 4
maps to dot1p priorities 5, 6, and 7.
Dell Networking OS Behavior: As soon as you apply a DCB policy with PFC enabled on an interface, DCBx starts exchanging
information with PFC-enabled peers. The IEEE802.1Qbb, CEE, and CIN versions of PFC Type, Length, Value (TLV) are supported.
DCBx also validates PFC congurations that are received in TLVs from peer devices.
NOTE: You cannot enable PFC and link-level ow control at the same time on an interface.
Conguring Lossless Queues
DCB also supports the manual conguration of lossless queues on an interface when PFC mode is turned o.
Prerequisite: A DCB with PFC conguration is applied to the interface with the following conditions:
• PFC mode is o (no pfc mode on).
• No PFC priority classes are congured (no pfc priority priority-range).
The conguration of no-drop queues provides exibility for ports on which PFC is not needed but lossless trac should egress from
the interface.
Data Center Bridging (DCB)
219