QoS with VLAN
12-14 Configuring Quality of Service
QoS with MLPPP multi-class regulates the output queue in such a way that, ideally, there is at
most one non-priority packet in front of the priority packet so the greatest latency that latency-
sensitive packets experience is never bigger than the fragment delay. Practically speaking, latency
for priority packets may be in the range of one to three fragment delays, depending on the traffic,
link speed and type of interface used.
QoS with MLPPP multi-class is configured by adding a class map for priority traffic and marking
it as a high priority class in the QoS policy map. That will allow priority packets to be scheduled
before other packets. Then to activate QoS, the policy is applied to the MLPPP interface.
Refer to “QoS with MLPPP Multi-Class Policy” on page12-26 for a configuration example.

Configuring QoS with FRF.12

Configuring QoS with FRF.12 is similar to configuring QoS with MLPPP multi-class except that
the FRF.12 fragment size is configured in bytes. The smaller the fragment size the smaller latency
for priority packets but the bigger the overhead for packet processing and link utilization.
Besides raising concerns about efficiency and utilization, setting a low fragment size may negate
the advantage of FRF.12 interleaving which interleaves packets within fragments. That is, multiple
fragmented packets are not interleaved. If the fragment size is smaller than the size of priority
packets, the priority packets will be chopped and FRF.12 will not insert its fragments into the
output fragme nt stream. In this case, fragments from priority pac ket must wait unt il all fragments
from the previous packets are transmitted thus causing heavy latency. In order to maximize the
benefit from FRF.12 interleaving, priority packets should be smaller than the fragment size. This
restriction does not exist for MLPPP multi-class.
For a sample configuration, refer to “QoS with FRF.12 Policy” on page12-27. Note that the same
QoS configuration is used for MLPPP and Frame Relay.
QoS with VLAN
The IEEE 802.1P signaling technique prioritizes network traffic at the data-link/MAC sublayer
(Layer 2). The 802.1P header includes a three-bit field for prioritization, which allows packets to be
grouped into various traffic classes. 802.1P is a spin-off of the 802.1Q (VLANs tagging) standard
and they work in tandem. The 802.1Q standard specifies a tag that appends to a MAC frame. The
two-part VLAN tag carries VLAN information including the 12-bit VLAN ID and 3-bit
Prioritization.
Additionally, 802.1P establishes eight levels of priority similar to IP Precedence levels. The XSR
permits mapping 802.1P prioritization to IP Precedence or DSCP before forwarding to the output
line and in reverse.

Traffic Classification

The XSR provides traffic classification based on the priority bits in the VLAN header, as illustrated
graphically below. For a VLAN connection, the XSR can use the 3-bit traffic class data in the 802.1q
header to convey the QoS classification of the encapsulated data.
Figure 12-2 Priority Information within VLAN 802.1q Header
Priority
802.1Q Tag Type CFI VLAN Identifier