SSH connection to secondary node get's timeout

MadDog2K · Jul 17, 2009, 7:35 PM

Hi,

I have a 2-node pfsense carp-setup, currently running 1.2.3 pre-release code from february 25th.
Reason was that I wondered if it could 'fix' the issue I'll describe below, I had with previous 1.2.x versions.
It apparantly does not, so I hope someone could shed some light on this :

4 nic's in use : 1 as sync, 1 as lan, 1 as wan and 1 as wan_backup (the latter only since recently)

The lan nic uses vlan's, and has carp for each vlan.

The lan interface has vlan tag 99
The first node has 10.128.99.2, the secondary node has 10.128.99.3 and their vip is 10.128.99.1

Workstations, servers, etc are in other vlans and subnets.
Whenever I have an SSH-session from a (mgmt-)workstation or server to the secondary node,
the connection get's a timeout after a while. Even when I have a ping open in it
(thus performing activity, and sending back that data to my SSH-client).

However, the connection to my first node never get's a timeout.

Could someone explain what is happening here ?

I have a feeling that somehow the 2nd node get's confused, because it has
an interface/IP in the same subnet as my workstation/server, but the connection is
coming via the first node.

I did not have this trouble when I was running native FreeBSD 5/6/7,
though I have to admit I did not run a firewall in between subnets back then (only on WAN->LAN, and vice-versa).

Anyhow : is this normal behaviour and is there a solution available ?
It's quite unhandy :)

Thanks !