Hi List,
I am running into a problem where I have 2 interfaces bridged with and ip address assigned.
I have another interface in which traffic has ingress traffic that needs to go out the bridged interface. I am trying unsuccessfully to SNAT the traffic leaving the bridge interface to its assigned address.
# brctl show xbrdg0 bridge name bridge id STP enabled interfaces xbrdg0 8000.000c297aa55f no eth0 eth1 # ip a s xbrdg0 11: xbrdg0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN link/ether 00:0c:29:7a:a5:5f brd ff:ff:ff:ff:ff:ff inet 192.168.100.3/24 scope global xbrdg0
# ip a s eth5 7: eth5: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP qlen 1000 link/ether 00:0c:29:7a:a5:7d brd ff:ff:ff:ff:ff:ff inet 10.10.0.1/29 scope global eth5
default via 192.168.100.1 dev xbrdg0
So I want traffic coming in eth5 with 10.10.0.x addresses to be source natted to 192.168.100.3. But my iptables nat statement never gets hit.
Chain POSTROUTING (policy ACCEPT 172 packets, 31384 bytes) pkts bytes target prot opt in out source destination 0 0 SNAT all -- * xbrdg0 0.0.0.0/0 0.0.0.0 to:192.168.100.3 29 1933 MASQUERADE all -- * tun+ 0.0.0.0/0 0.0.0.0/0
# ping -I 10.10.0.1 8.8.8.8
# tcpdump -nli xbrdg0 icmp or arp tcpdump: verbose output suppressed, use -v or -vv for full protocol decode listening on xbrdg0, link-type EN10MB (Ethernet), capture size 65535 bytes 12:52:06.914295 IP 10.10.0.1 > 8.8.8.8: ICMP echo request, id 38932, seq 1, length 64 12:52:07.914592 IP 10.10.0.1 > 8.8.8.8: ICMP echo request, id 38932, seq 2, length 64 12:52:08.914579 IP 10.10.0.1 > 8.8.8.8: ICMP echo request, id 38932, seq 3, length 64
Any ideas?
Thanks, Steve
On 01/20/2016 04:21 PM, Gordon Messmer wrote:
On 01/20/2016 09:55 AM, Steve Clark wrote:
Any ideas?
IP forwarding needs to be enabled, and you also need rules in your FORWARD chain to allow the packets.
Thanks, but forwarding is turned on and my FW rules are empty.
Chain INPUT (policy ACCEPT 359K packets, 136M bytes) pkts bytes target prot opt in out source destination
Chain FORWARD (policy ACCEPT 55801 packets, 4736K bytes) pkts bytes target prot opt in out source destination
Chain OUTPUT (policy ACCEPT 319K packets, 141M bytes) pkts bytes target prot opt in out source destination
On 01/21/2016 03:49 AM, Steve Clark wrote:
Thanks, but forwarding is turned on and my FW rules are empty.
Try specifying the physical device the packets are going out, rather than the bridge, in your postrouting rule.
Apparently you also need an ebtables rule to prevent the return packets from being merely bridged? http://serverfault.com/questions/349688/iptables-bridge-nat-setup
I'd test it, but have a look if you get the outbound traffic working and return traffic doesn't.
On 20/01/2016 19:55, Steve Clark wrote:
So I want traffic coming in eth5 with 10.10.0.x addresses to be source natted to 192.168.100.3. But my iptables nat statement never gets hit.
Chain POSTROUTING (policy ACCEPT 172 packets, 31384 bytes) pkts bytes target prot opt in out source destination 0 0 SNAT all -- * xbrdg0 0.0.0.0/0 0.0.0.0 to:192.168.100.3 29 1933 MASQUERADE all -- * tun+ 0.0.0.0/0 0.0.0.0/0
# ping -I 10.10.0.1 8.8.8.8
First you should try to match without SNAT at all with a simple log target and see if it matches. I would start with: iptables -t nat -I POSTROUTING -s 10.0.0.1 -o xbrdg0 -j LOG --log-prefix "Should-SNAT: " --log-level 4
And then: iptables -t nat -I POSTROUTING -s 10.0.0.1 -o xbrdg0 -j SNAT --to-source 192.168.100.3
And see what happens. Also there might be something about this bridge settings and it maybe needs the "-o eth1" but it would be a bit weird.
Eliezer
-
Eliezer Croitoru -
Gordon Messmer -
Steve Clark