carp and arp_rtrequest: bad gateway 1.1.1.5 (!AF_LINK)

[c0re]

Further workaround:

1st server has advbase 1 and advskew 0
2nd server has advbase 1 and advskew 100

So 2nd server should failover after what period of time when 1st server fails?

In openbsd man 8 ifconfig (
http://www.openbsd.org/cgi-bin/man.cgi?query=ifconfig&apropos=0&sektion=8&manpath=OpenBSD+Current&arch=i386&format=html
) I found that info:


     Taken together, the advbase and advskew indicate how frequently, in
     seconds, the host will advertise the fact that it considers itself master
     of the virtual host.  The formula is advbase + (advskew / 256).  If the
     master does not advertise within three times this interval, this host
     will begin advertising as master.

Due to CARP be ported from openbsd I think it should be same for freebsd too.
So 2nd server should take MASTER state after 3*(advbase+(advskew /
256)) = 3*(1+(100/256)) =~ 4 secs.
But when I promote 1st servers CARP interface down, 2nd server became
MASTER immediately, no 4 seconds holdtime/timeout occurs.
Why is it so?
Any tips?
Want to understand why 2nd server takes master state too often while
1st server are still available in network.