just to make sure: changing the sshd configuration fixes your problem as
well? i.e. there is no bug?
Putting ListenAddress $IP rdomain 255 inside sshd_config did the trick,
netstat -nr -f inet was looking at it's default route table which was
255 rtable in this case.
Running sshd_rtable=255 I
Thinking about it a little more, I've configured sshd_rtable=255 on
/etc/rc.conf.local,
on the non-working boxes, I'm re-checking this doing:
rcctl set sshd rtable 0
and changing sshd_config:
ListenAddress $IP rdomain 255
Thank you
On 4/12/18 12:04, Nicolas Pence wrote:
I
I'm using 3 different rdomains, with one BGPd instance in each of them
with different configurations, when using rdomain 255 some routing info
is leaked into rtable 0 to the point that doesn't allow to route, this
happens over a few minutes/hours of functioning (tested on two VM's with
6.2-stab
