FD.io VPP
v21.01.1
Vector Packet Processing
|
set urpf [ip4|ip6] [rx|tx] [off|strict|loose] <INTERFACE>.
This command configures uRPF on an interface. Two flavours are supported (the default is strict):
Example of graph node before range checking is enabled:
vpp# show vlib graph ip4-rx-urpf-strict
Name Next Previous ip4-rx-urpf-strict ip4-drop [0]
Example of how to enable unicast source checking on an interface:
vpp# set urpf ip4 rx GigabitEthernet2/0/0 loose
Example of graph node after range checking is enabled:
vpp# show vlib graph ip4-rx-urpf-loose
Name Next Previous ip4-rx-urpf-loose ip4-drop [0] ip4-input-no-checksum ip4-source-and-port-range- ip4-input
Example of how to display the feature enabed on an interface:
vpp# show ip interface features GigabitEthernet2/0/0
IP feature paths configured on GigabitEthernet2/0/0...
ipv4 unicast: ip4-rx-urpf-loose ip4-lookup
ipv4 multicast: ip4-lookup-multicast
ipv4 multicast: interface-output
ipv6 unicast: ip6-lookup
ipv6 multicast: ip6-lookup
ipv6 multicast: interface-output
Example of how to disable unicast source checking on an interface:
vpp# set urpf ip4 off GigabitEthernet2/0/0
Declaration: set_interface_ip_source_check_command (src/plugins/urpf/urpf.c line 233)
Implementation: urpf_cli_update.
urpf-accept [table <table-id>] [add|del] <PREFIX>.
Add an exemption for a prefix to pass the Unicast Reverse Path Forwarding (uRPF) loose check. This is for testing purposes only. If the '<em>table</em>' is not enter it is defaulted to 0. Default is to '<em>add</em>'. VPP always performs a loose uRPF check for for-us traffic.
vpp# set urpf-accept table 7 10.0.0.0/8 add
Declaration: urpf_accept_command (src/plugins/urpf/urpf.c line 310)
Implementation: urpf_cli_accept.