|
From: T S. <tsu...@no...> - 2006-03-23 12:09:08
|
> NetBSD does more things on kernel about ports, and it probably solves
> at least some of the problems.
>=20
> I'll have to make a test configuration between all three
> implementations to see exactly how they deal with such problems.
Yvan, thanks. That would be great.=20
I just checked the code of NetBSD. The code for previously mentioned case =
is same. So, I think the patch is not there. Though, I have not tested the =
scenario with the setup. If somebody has a netbsd peer-peer racoon without =
NAT support, please let me know SIGHUP at one peer sends a ipsec-sa delete =
message to other. =20
Another question, when an ipsec-sa delete message is received, should =
racoon delete the intbound ipsec-sa too? Rightnow, there is a comment =
saying in purge_ipsec_spi (isakmp_inf.c:1178).
/* don't delete inbound SAs at the moment */
/* XXX should we remove SAs with opposite direction as well? */
suresh.
|