Table of Contents
Advertisement
About the PoS Modules
Service Provider Features
The PoS modules provide the following features for service provider environments:
• DSCP mapping
• VLAN ID (VID) tag mapping
• VLAN ID (VID) tag nesting
• Port tunneling
You can use the
command to configure a mapped relationship
diffserv dscp-mapping
between an input DSCP and an associated output DSCP. Each PoS port supports three
DSCP mapping tables: one of the tables is used in the ingress direction; two are used for
egress flows (onto the SONET link). The two egress tables are for the congested and
noncongested states, as determined by the RED algorithm. If RED is not enabled on the
PoS port, the egress congested-state mapping table is not used.
In the ingress direction, the input DSCP of a packet received from the SONET link is
replaced by an output DSCP before the packet is forwarded. In the egress direction, the
operation is similar, except that the DSCP mapping occurs before the packet is
transmitted onto the SONET link.
One potential use of the DSCP mapping capability is to reconcile varying DiffServ
policies at the boundary between autonomous systems, such as at the boundary
between two ISPs. The availability of different tables for the congested and
noncongested states is useful in marking operations that increase the probability of
packets being dropped during times of congestion, as discussed in the DiffServ Assured
Forwarding RFC (RFC 2597).
An analogous feature has been added for managing 802.1Q tags. The
command provides support for VLAN ID (VID) mapping tables.
dot1q tagmapping
Each PoS port supports two VID tables: one table is used in the ingress direction; the
other is used in the egress direction. Each of the tables enables an input VID to be
mapped to an output VID. This feature is useful in reconciling policy differences at the
boundary between the customer and the service provider.
Another related enhancement provides support for nested 802.1Q tags by allowing a
tag push or tag pop attribute to be associated with a VID. The push attribute indicates that
a new tag is to be added to the frame, while the pop attribute indicates that the top-level
tag is to be removed from the frame. This capability is augmented by an option that
allows the 802.1p priority of the frame to be either preserved or set to a
user-configurable value when a new tag is pushed. These functions make it possible for
service providers to tunnel customer-specific VLANs across a common SONET
backbone in a very simple manner.
Packet Over SONET Module Installation and User Guide
1-11
Advertisement
Table of Contents
