36-95
Catalyst 3560 Switch Software Configuration Guide
OL-8553-06
Chapter 36 Configuring IP Unicast Routing
Configuring Protocol-Independent Features
You can apply a policy route map to an EtherChannel port channel in Layer 3 mode, but you cannot
apply a policy route map to a physical interface that is a member of the EtherChannel. If you try to
do so, the command is rejected. When a policy route map is applied to a physical interface, that
interface cannot become a member of an EtherChannel.
You can define a maximum of 246 IP policy route maps on the switch.
You can define a maximum of 512 access control entries (ACEs) for PBR on the switch.
When configuring match criteria in a route map, follow these guidelines:
Do not match ACLs that permit packets destined for a local address. PBR would forward these
packets, which could cause ping or Telnet failure or route protocol flapping.
Do not match ACLs with deny ACEs. Packets that match a deny ACE are sent to the CPU, which
could cause high CPU utilization.
To use PBR, you must first enable the routing template by using the sdm prefer routing global
configuration command. PBR is not supported with the VLAN or default template. For more
information on the SDM templates, see Chapter 8, “Configuring SDM Templates.
VRF and PBR are mutually exclusive on a switch interface. You cannot enable VRF when PBR is
enabled on an interface. The reverse is also true, you cannot enable PBR when VRF is enabled on
an interface.
Web Cache Communication Protocol (WCCP) and PBR are mutually exclusive on a switch
interface. You cannot enable WCCP when PBR is enabled on an interface. The reverse is also true,
you cannot enable PBR when WCCP is enabled on an interface.
The number of TCAM entries used by PBR depends on the route map itself, the ACLs used, and the
order of the ACLs and route-map entries.
Policy-based routing based on packet length, TOS, set interface, set default next hop, or set default
interface are not supported. Policy maps with no valid set actions or with set action set to Don’t
Fragme nt are not supported.
The switch supports quality of service (QoS) DSCP and IP precedence matching in PBR route maps,
with these limitations:
You cannot apply QoS DSCP mutation maps and PBR route maps to the same interface.
You cannot configure DSCP transparency and PBR DSCP route maps on the same switch.
When you configure PBR with QoS DSCP, you can set QoS to be enabled (by entering the mls
qos global configuration command) or disabled (by entering the no mls qos command). When
QoS is enabled, to ensure that the DSCP value of the traffic is unchanged, you should configure
a DSCP trust state on the port where traffic enters the switch by entering the mls qos trust dscp
interface configuration command. If the trust state is not DSCP, by default all nontrusted traffic
would have the DSCP value marked as 0.
Enabling PBR
By default, PBR is disabled on the switch. To enable PBR, you must create a route map that specifies
the match criteria and the resulting action if all of the match clauses are met. Then, you must enable PBR
for that route map on an interface. All packets arriving on the specified interface matching the match
clauses are subject to PBR.
PBR can be fast-switched or implemented at speeds that do not slow down the switch. Fast-switched
PBR supports most match and set commands. PBR must be enabled before you enable fast-switched
PBR. Fast-switched PBR is disabled by default.