Discussion:
Bridge with 1x eth and 2x tun, adding 2nd eth, hosts become unavailable
(too old to reply)
Marc Roos
2018-04-07 07:59:37 UTC
Permalink
In this situation I can ping a vm from br0 connected to 172.16.1.y



+-----------+
| BR0 |
172.16.1.y | | 172.16.1.x
--------|ETH2 TUN1|-------
| |
| | 172.16.1.x
| TUN2|-------
| |
+-----------+

When only adding a 2nd eth adapter with 'brctl addif br0 eth1' connected
to 10.11.12.y. The same ping is not working anymore.


+-----------+
| BR0 |
172.16.1.y | | 172.16.1.x
--------|ETH2 TUN1|-------
| |
| | 172.16.1.x
10.11.12.y | TUN2|-------
--------|ETH1 |
+-----------+

When I remove the eth1 again, the ping starts working again.

Can someone explain me why the bridge is behaving like this? And how
this can be fixed?
Stephen Hemminger
2018-04-08 16:26:21 UTC
Permalink
On Sat, 7 Apr 2018 09:59:37 +0200
Post by Marc Roos
In this situation I can ping a vm from br0 connected to 172.16.1.y
+-----------+
| BR0 |
172.16.1.y | | 172.16.1.x
--------|ETH2 TUN1|-------
| |
| | 172.16.1.x
| TUN2|-------
| |
+-----------+
When only adding a 2nd eth adapter with 'brctl addif br0 eth1' connected
to 10.11.12.y. The same ping is not working anymore.
+-----------+
| BR0 |
172.16.1.y | | 172.16.1.x
--------|ETH2 TUN1|-------
| |
| | 172.16.1.x
10.11.12.y | TUN2|-------
--------|ETH1 |
+-----------+
When I remove the eth1 again, the ping starts working again.
Can someone explain me why the bridge is behaving like this? And how
this can be fixed?
Are you creating a spanning tree loop? Also. Bridge is a Layer 2 device
not a Layer 3 device; therefore generally all interfaces in a bridge have same
subnet.
Marc Roos
2018-04-08 16:12:48 UTC
Permalink
Post by Stephen Hemminger
Post by Marc Roos
In this situation I can ping a vm from br0 connected to 172.16.1.y
+-----------+
| BR0 |
172.16.1.y | | 172.16.1.x
--------|ETH2 TUN1|-------
| |
| | 172.16.1.x
| TUN2|-------
| |
+-----------+
When only adding a 2nd eth adapter with 'brctl addif br0 eth1'
connected
Post by Stephen Hemminger
Post by Marc Roos
to 10.11.12.y. The same ping is not working anymore.
+-----------+
| BR0 |
172.16.1.y | | 172.16.1.x
--------|ETH2 TUN1|-------
| |
| | 172.16.1.x
10.11.12.y | TUN2|-------
--------|ETH1 |
+-----------+
When I remove the eth1 again, the ping starts working again.
Can someone explain me why the bridge is behaving like this? And how
this can be fixed?
Are you creating a spanning tree loop? Also. Bridge is a Layer 2
device
Post by Stephen Hemminger
not a Layer 3 device; therefore generally all interfaces in a bridge
have same
Post by Stephen Hemminger
subnet.
I don’t think I am creating a loop, I have stp on and ebtables:
-i eth2 -o eth1 -j DROP , pcnt = 301 -- bcnt = 17100
-i eth1 -o eth2 -j DROP , pcnt = 3787 -- bcnt = 180148

I would like to achieve that what ever hosts and networks I connect on
ethx on tunx, could see the other ones located on the the other ethx or
tunx. Is this not possible with the bridge? Is there another technique
to do this?

Continue reading on narkive:
Loading...