3 tincds, one down, pinging down machine from others loses connection

[ebina1]

I have a VPN (192.168.0.0/16) with three subnets each gatewayed by tincd
call them A, B, and C.

C:(192.168.0.0/24), A:(192.168.1.0/24), and B:(192.168.2.0/24)

ConnectTo are set up
A->C, A->B
B->C

When all three machines are up, everything works fine.
If B goes down, everything is still fine.  tincd with debug
shows periodic attempts for A to contact B, but A and C still
talk.  If from A you attempt to ping B (while it is down)
A stops talking to C.  No error messages in the log, and A
still periodically tries to contact B, but it acts like C
just isn't there.

C notices A not responding, and closes connection to C,
but C never attempts to reconnect.

tinc.conf on C:
Name = C
PrivateKeyFile = /usr/local/etc/tinc/krmvpn/rsa_key.priv
TapDevice = /dev/tap0
KeyExpire = 30000000

tinc.conf on A:
Name = A
PrivateKeyFile = /usr/local/etc/tinc/krmvpn/rsa_key.priv
TapDevice = /dev/tun
KeyExpire = 30000000




Tinc:         Discussion list about the tinc VPN daemon
Archive:      http://mail.nl.linux.org/lists/
Tinc site:    http://tinc.nl.linux.org/