
Chapter 67 Configuring IP Filter Rules 635
BCM 4.0 Networking Configuration Guide
Differentiated Services (DiffServ) overview
Differentiated services (DiffServ) is a Quality of Service (QoS) network architecture that offers
varied levels of service for different types of data traffic. DiffServ allows you to designate a
specific level of performance on a packet-by-packet basis instead of using the “best-effort” model
for your data delivery. You can give preferential treatment (prioritization) to applications that
require high performance and reliable service, such as voice and video over IP.
BCM includes the capability to enhance your network traffic management. For each packet, there
is an octet in the packet header, the DiffServ (DS) field, that you can designate for specific service.
For IP packets, six bits of the DiffServ field is the DiffServ Code Point (DSCP). The DSCP value
defines how the packet is to be treated as it travels through the network. You can set traffic criteria
to match the DS field, and rule actions to change the DiffServ field to conform to various other
mappings.
DiffServ IP Quality of Service (QoS) architecture
DiffServ uses a simple mechanism that relies on a special encoding of the first 6 bits of the
DiffServ byte in the IP header. This byte is the IPv4 Type of Service (ToS) byte; for IPv6, is the
Traffic Class byte. The first 6 bits of this byte are called the DiffServ Code Point (DSCP).
In the packet forwarding path, differentiated services are processed by mapping the packet DSCP
to a particular forwarding treatment, or per hop behavior (PHB), at each network node along its
path. The code points may be chosen from a set of 32 standard values, a set of 16 recommended
values to be used in the future, or a set of 16 values reserved for experimentation and local use. Of
the 32 standard values, there are 8 Class Selector code points that are used primarily (but not
exclusively) for backward compatibility with existing definitions of the ToS byte.
BCM is a DiffServ node that can support DiffServ functions and behavior. DiffServ architecture
defines a DiffServ-capable domain as a contiguous set of DiffServ-compliant nodes that operate
with a common set of service provisioning policies and PHB definitions. The DiffServ domain is
an autonomous system or network such as an internet service provider (ISP) network or campus
LAN.
DiffServ assumes the existence of a service level agreement (SLA) between DiffServ domains that
share a border. The SLA defines the profile for the aggregate traffic flowing from one network to
the other based on rule criteria. In a given traffic direction, the traffic is expected to be shaped at
the egress point of the upstream network and policed at the ingress point of the downstream
network.
End-to-end QoS is enabled, typically through bilateral agreements (an agreement between two
DiffServ domains), between all the domains from the sender to the receiver. These agreements aid
in consistent PHB and QoS performance across all domains.
Typically, there are three types of edge devices in a DiffServ domain:
• Edge node (EN) — the switch or router connected directly to the desktop end station (ES)
(BCM is an edge node in the DiffServ domain)
• Ingress border node (IBN) — the ingress router at the boundary between two DiffServ
domains
• Egress border node (EBN) — the egress router at the boundary between two DiffServ domains
Kommentare zu diesen Handbüchern