When configuring VRRP Aware PIM, consider the following recommendations.
VRRP could be configured to use BFD to speed up failure detection in addition to the functionality provided by VRRP Aware PIM.
To optimize failover, the config>router>pim>non-dr-attract-traffic command can be enabled on the primary and secondary routers to make them a hot-standby redundant pair. This configuration ignores the DR state and attracts traffic to populate the router’s PIM database. This setting should not be used if multicast traffic must only follow the VRRP MR.
The config>service>oper-group>hold-time>group>up time on the primary router and config>service>oper-group>hold-time>group>down time on the secondary router should both be set to the time needed to repopulate the PIM database; for example, 10 seconds. This allows the primary router to populate its PIM database again before becoming the DR if a failure occurs from the primary to secondary router, and recover from the secondary back to the primary router.
The config>service>oper-group>hold-time>group>up time should be set to 0 on the secondary router so that it assumes the DR role immediately if the primary router fails. The up hold time is set to 4 seconds by default, which delays the DR change unnecessarily.
The sticky DR setting should be disabled if it is configured with the config>router>pim>if>sticky-dr command. Sticky DR enables the secondary router to continue to act as the DR after the primary router comes back up. Sticky DR is incompatible with the VRRP Aware PIM mechanism that tracks the VRRP MR.
The following is a basic configuration example for VRRP Aware PIM.
service
oper-group ‟VAwP1” create
exit
vprn 1 customer 1 create
interface to-LAN
vrrp 1 create
oper-group ‟VAwP1”
exit
pim
interface to-LAN
monitor-oper-group ‟VAwP1” family ipv4 add 90
monitor-oper-group ‟VAwP1” family ipv6 add 90
exit
interface to-LAN2
monitor-oper-group ‟VAwP1” family ipv4 add 90
monitor-oper-group ‟VAwP2” family ipv6 set 90
exit
exit
exit