Concept Guide
• All 802.1p priorities are enabled for PFC. Queues to which PFC priority trac is mapped are lossless by default. Trac may be
interrupted due to an interface ap (going down and coming up).
• For PFC to be applied on an Aggregator port, the auto-congured priority trac must be supported by a PFC peer (as detected by
DCBx).
• A dcb-map for PFC applied to an interface may become invalid if dot1p-queue mapping is recongured. This situation occurs when the
new dot1p-queue assignment exceeds the maximum number (2) of lossless queues supported globally on the switch. In this case, all
PFC congurations received from PFC-enabled peers are removed and re-synchronized with the peer devices.
• Dell Networking OS does not support MACsec Bypass Capability (MBC).
Conguring Enhanced Transmission Selection
ETS provides a way to optimize bandwidth allocation to outbound 802.1p classes of converged Ethernet trac.
Dierent trac types have dierent service needs. Using ETS, you can create groups within an 802.1p priority class to congure dierent
treatment for trac with dierent bandwidth, latency, and best-eort needs.
For example, storage trac is sensitive to frame loss; interprocess communication (IPC) trac is latency-sensitive. ETS allows dierent
trac types to coexist without interruption in the same converged link by:
• Allocating a guaranteed share of bandwidth to each priority group.
• Allowing each group to exceed its minimum guaranteed bandwidth if another group is not fully using its allotted bandwidth.
To congure ETS and apply an ETS output policy to an interface, you must:
1 Create a Quality of Service (QoS) output policy with ETS scheduling and bandwidth allocation settings.
2 Create a priority group of 802.1p trac classes.
3 Congure a DCB output policy in which you associate a priority group with a QoS ETS output policy.
4 Apply the DCB output policy to an interface.
How Enhanced Transmission Selection is Implemented
Enhanced transmission selection (ETS) provides a way to optimize bandwidth allocation to outbound 802.1p classes of converged Ethernet
trac. Dierent trac types have dierent service needs. Using ETS, groups within an 802.1p priority class are auto-congured to provide
dierent treatment for trac with dierent bandwidth, latency, and best-eort needs.
For example, storage trac is sensitive to frame loss; interprocess communication (IPC) trac is latency-sensitive. ETS allows dierent
trac types to coexist without interruption in the same converged link.
NOTE
: The IEEE 802.1Qaz, CEE, and CIN versions of ETS are supported.
ETS is implemented on an Aggregator as follows:
• Trac in priority groups is assigned to strict-queue or WERR scheduling in a dcb-map and is managed using the ETS bandwidth-
assignment algorithm. Dell Networking OS de-qeues all frames of strict-priority trac before servicing any other queues. A queue with
strict-priority trac can starve other queues in the same port.
• ETS-assigned bandwidth allocation and scheduling apply only to data queues, not to control queues.
• Dell Networking OS supports hierarchical scheduling on an interface. Dell Networking OS control trac is redirected to control queues
as higher priority trac with strict priority scheduling. After control queues drain out, the remaining data trac is scheduled to queues
according to the bandwidth and scheduler conguration in the dcb-map. The available bandwidth calculated by the ETS algorithm is
equal to the link bandwidth after scheduling non-ETS higher-priority trac.
• By default, equal bandwidth is assigned to each port queue and each dot1p priority in a priority group.
• By default, equal bandwidth is assigned to each priority group in the dcb-map applied to an egress port. The sum of auto-congured
bandwidth allocation to dot1p priority trac in all ETS priority groups is 100%.
• dot1p priority trac on the switch is scheduled according to the default dot1p-queue mapping. dot1p priorities within the same queue
should have the same trac properties and scheduling method.
• A priority group consists of 802.1p priority values that are grouped together for similar bandwidth allocation and scheduling, and that
share the same latency and loss requirements. All 802.1p priorities mapped to the same queue should be in the same priority group.
50
Data Center Bridging (DCB)