Service Manual

Unused bandwidth
usage:
Normally, if there is no trac or unused bandwidth for a priority group, the bandwidth allocated to the group
is distributed to the other priority groups according to the bandwidth percentage allocated to each group.
However, when three priority groups with dierent bandwidth allocations are used on an interface:
• If priority group 3 has free bandwidth, it is distributed as follows: 20% of the free bandwidth to priority
group 1 and 30% of the free bandwidth to priority group 2.
• If priority group 1 or 2 has free bandwidth, (20 + 30)% of the free bandwidth is distributed to priority
group 3. Priority groups 1 and 2 retain whatever free bandwidth remains up to the (20+ 30)%.
Strict-priority
groups:
If two priority groups have strict-priority scheduling, trac assigned from the priority group with the higher
priority-queue number is scheduled rst. However, when three priority groups are used and two groups have
strict-priority scheduling (such as groups 1 and 3 in the example), the strict priority group whose trac is
mapped to one queue takes precedence over the strict priority group whose trac is mapped to two
queues.
Therefore, in this example, scheduling trac to priority group 1 (mapped to one strict-priority queue) takes precedence over
scheduling trac to priority group 3 (mapped to two strict-priority queues).
Priority-Based Flow Control Using Dynamic Buer Method
Priority-based ow control using dynamic buer spaces is supported on the platform.
In a data center network, priority-based ow control (PFC) manages large bursts of one trac type in multiprotocol links so that it
does not aect other trac types and no frames are lost due to congestion. When PFC detects congestion on a queue for a
specied priority, it sends a pause frame for the 802.1p priority trac to the transmitting device.
Pause and Resume of Trac
The pause message is used by the sending device to inform the receiving device about a congested, heavily-loaded trac state that
has been identied. When the interface of a sending device transmits a pause frame, the recipient acknowledges this frame by
temporarily halting the transmission of data packets. The sending device requests the recipient to restart the transmission of data
trac when the congestion eases and reduces. The time period that is specied in the pause frame denes the duration for which
the ow of data packets is halted. When the time period elapses, the transmission restarts.
When a device sends a pause frame to another device, the time for which the sending of packets from the other device must be
stopped is contained in the pause frame. The device that sent the pause frame empties the buer to be less than the threshold
value and restarts the acceptance of data packets.
Dynamic ingress buering enables the sending of pause frames at dierent thresholds based on the number of ports that experience
congestion at a time. This behavior impacts the total buer size used by a particular lossless priority on an interface. The pause and
resume thresholds can also be congured dynamically. You can congure a buer size, pause threshold, ingress shared threshold
weight, and resume threshold to control and manage the total amount of buers that are to be used in your network environment.
Buer Sizes for Lossless or PFC Packets
You can congure up to a maximum of 4 lossless (PFC) queues. By conguring 4 lossless queues, you can congure 4 dierent
priorities and assign a particular priority to each application that your network is used to process. For example, you can assign a
higher priority for time-sensitive applications and a lower priority for other services, such as le transfers. You can congure the
amount of buer space to be allocated for each priority and the pause or resume thresholds for the buer. This method of
conguration enables you to eectively manage and administer the behavior of lossless queues.
Although the system contains of space for shared buers, a minimum guaranteed buer is provided to all the internal and external
ports in the system for both unicast and multicast trac. This minimum guaranteed buer reduces the total available shared buer
to . This shared buer can be used for lossy and lossless trac.
Data Center Bridging (DCB)
245