I think it's just old mistake propagated through history. I've in many networks configured default ARP timeout below 300s to fix this problem. The default behaviour imho is just broken, it can even cause loops in sane topologies/configurations.
On 11 December 2017 at 18:27, Karl Gerhard <[email protected]> wrote: > Hello > > by default > ARP table entries will be saved for 20 minutes and > MAC table entries will be saved for 5minutes. > > This seems to be a rather stupid default to me since expired MAC table > entries (in conjuction with still existing ARP table entries) will cause > Unknown Unicast packets to be flooded. We've been bitten by this because we > did VRRP between two routers and two switches and traffic flow was asymmetric > so one switch forgot the MAC addresses and flooded Unknown Unicasts. > > Is there a logic behind those defaults or am I right to assume that MAC table > timers should /always/ be higher than ARP table timers? > > Links to the docs: > https://www.juniper.net/documentation/en_US/junos/topics/task/configuration/configuring-arp-aging-timer.html > https://www.juniper.net/documentation/en_US/junos/topics/reference/configuration-statement/mac-table-aging-time-bridging.html > > Regards > Karl > > _______________________________________________ > juniper-nsp mailing list [email protected] > https://puck.nether.net/mailman/listinfo/juniper-nsp -- ++ytti _______________________________________________ juniper-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/juniper-nsp
