Reference Guide

Applying Egress Layer 3 ACLs (Control-Plane)
By default, packets originated from the system are not filtered by egress ACLs.
For example, if you initiate a ping session from the system and apply an egress ACL to block this type of traffic on the
interface, the ACL does not affect that ping traffic. The Control Plane Egress Layer 3 ACL feature enhances IP
reachability debugging by implementing control-plane ACLs for CPU-generated and CPU-forwarded traffic. Using permit
rules with the count option, you can track on a per-flow basis whether CPU-generated and CPU-forwarded packets
were transmitted successfully.
NOTE: The ip control-plane [egress filter] and the ipv6 control-plane [egress
filter]
commands are not supported on the S4820T system.
1. Apply Egress ACLs to IPv4 system traffic.
CONFIGURATION mode
ip control-plane [egress filter]
2. Apply Egress ACLs to IPv6 system traffic.
CONFIGURATION mode
ipv6 control-plane [egress filter]
3. Create a Layer 3 ACL using permit rules with the count option to describe the desired CPU traffic.
CONFIG-NACL mode
permit ip {source mask | any | host ip-address} {destination mask | any |
host ip-address} count
FTOS Behavior: Virtual router redundancy protocol (VRRP) hellos and internet group management protocol (IGMP)
packets are not affected when you enable egress ACL filtering for CPU traffic. Packets sent by the CPU with the source
address as the VRRP virtual IP address have the interface MAC address instead of VRRP virtual MAC address.
IP Prefix Lists
Prefix lists are supported on the S4820T platform.
IP prefix lists control routing policy. An IP prefix list is a series of sequential filters that contain a matching criterion
(examine IP route prefix) and an action (permit or deny) to process routes. The filters are processed in sequence so that
if a route prefix does not match the criterion in the first filter, the second filter (if configured) is applied. When the route
prefix matches a filter, FTOS drops or forwards the packet based on the filter’s designated action. If the route prefix does
not match any of the filters in the prefix list, the route is dropped (that is, implicit deny).
A route prefix is an IP address pattern that matches on bits within the IP address. The format of a route prefix is
A.B.C.D/X where A.B.C.D is a dotted-decimal address and /X is the number of bits that should be matched of the dotted
decimal address. For example, in 112.24.0.0/16, the first 16 bits of the address 112.24.0.0 match all addresses between
112.24.0.0 to 112.24.255.255.
The following examples show permit or deny filters for specific routes using the le and ge parameters, where x.x.x.x/x
represents a route prefix:
• To deny only /8 prefixes, enter deny x.x.x.x/x ge 8 le 8.
• To permit routes with the mask greater than /8 but less than /12, enter permit x.x.x.x/x ge 8.
• To deny routes with a mask less than /24, enter deny x.x.x.x/x le 24.
• To permit routes with a mask greater than /20, enter permit x.x.x.x/x ge 20.
The following rules apply to prefix lists:
109