Actions
Bug #173
closedMesh packets on bat0
Start date:
07/13/2013
Due date:
% Done:
0%
Estimated time:
Description
Batman 2013.2.0 sends incoming mesh packets to bat0 on Linux 2.6.32
Testsetup
Two indentical nodes, running Debian 6.0.7 (x86), eth1 of both nodes connected
node1 ~ # uname -a
Linux node1 2.6.32-5-686 #1 SMP Fri May 10 08:33:48 UTC 2013 i686 GNU/Linux
node1 ~ # dpkg -l | grep 2.6.32-5-686
ii linux-headers-2.6.32-5-686 2.6.32-48squeeze3 Header files for Linux 2.6.32-5-686
ii linux-image-2.6.32-5-686 2.6.32-48squeeze3 Linux 2.6.32 for modern PCs
node1 ~ # brctl show
bridge name bridge id STP enabled interfaces
node1 ~ # ip link set eth1 up
node1 ~ # batctl if add eth1
node1 ~ # batctl if
eth1: active
node1 ~ # ip link set bat0 up
node1 ~ # ip add add 10.10.10.1/24 dev bat0
node1 ~ # ip add
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 08:00:27:4b:75:b3 brd ff:ff:ff:ff:ff:ff
inet 10.1.1.215/24 brd 10.1.1.255 scope global eth0
inet6 fe80::a00:27ff:fe4b:75b3/64 scope link
valid_lft forever preferred_lft forever
3: eth1: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master bat0 state UP qlen 1000
link/ether 08:00:27:00:00:01 brd ff:ff:ff:ff:ff:ff
inet6 fe80::a00:27ff:fe00:1/64 scope link
valid_lft forever preferred_lft forever
4: bat0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN
link/ether 9a:fa:b3:73:f6:d8 brd ff:ff:ff:ff:ff:ff
inet 10.10.10.1/24 scope global bat0
inet6 fe80::98fa:b3ff:fe73:f6d8/64 scope link
valid_lft forever preferred_lft forever
node2 ~ # brctl show
bridge name bridge id STP enabled interfaces
node2 ~ # ip link set eth1 up
node2 ~ # batctl if add eth1
node2 ~ # batctl if
eth1: active
node2 ~ # ip link set bat0 up
node2 ~ # ip add add 10.10.10.2/24 dev bat0
node2 ~ # ip add
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 08:00:27:d9:50:55 brd ff:ff:ff:ff:ff:ff
inet 10.1.1.216/24 brd 10.1.1.255 scope global eth0
inet6 fe80::a00:27ff:fed9:5055/64 scope link
valid_lft forever preferred_lft forever
3: eth1: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast master bat0 state UP qlen 1000
link/ether 08:00:27:00:00:02 brd ff:ff:ff:ff:ff:ff
inet6 fe80::a00:27ff:fe00:2/64 scope link
valid_lft forever preferred_lft forever
4: bat0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UNKNOWN
link/ether b2:40:25:42:d3:05 brd ff:ff:ff:ff:ff:ff
inet 10.10.10.2/24 scope global bat0
inet6 fe80::b040:25ff:fe42:d305/64 scope link
valid_lft forever preferred_lft forever
node1 ~ # batctl o
[B.A.T.M.A.N. adv 2013.2.0, MainIF/MAC: eth1/08:00:27:00:00:01 (bat0)]
Originator last-seen (#/255) Nexthop [outgoingIF]: Potential nexthops ...
08:00:27:00:00:02 0.780s (255) 08:00:27:00:00:02 [ eth1]: 08:00:27:00:00:02 (255)
node2 ~ # batctl o
[B.A.T.M.A.N. adv 2013.2.0, MainIF/MAC: eth1/08:00:27:00:00:02 (bat0)]
Originator last-seen (#/255) Nexthop [outgoingIF]: Potential nexthops ...
08:00:27:00:00:01 0.420s (255) 08:00:27:00:00:01 [ eth1]: 08:00:27:00:00:01 (255)
Let's send some traffic to the mesh:
node1 ~ # ping 10.10.10.2 PING 10.10.10.2 (10.10.10.2) 56(84) bytes of data. 64 bytes from 10.10.10.2: icmp_req=1 ttl=64 time=1.97 ms 64 bytes from 10.10.10.2: icmp_req=2 ttl=64 time=0.318 ms 64 bytes from 10.10.10.2: icmp_req=3 ttl=64 time=0.314 ms [...]
On eth1 of node2 everything looks fine:
node2 ~ # batctl td eth1 23:55:17.975137 BAT 08:00:27:00:00:01 > 08:00:27:00:00:02: UCAST, ttvn 1, ttl 50, IP 10.10.10.1 > 10.10.10.2: ICMP echo request, id 1588, seq 34, length 64 23:55:17.975261 BAT 08:00:27:00:00:02 > 08:00:27:00:00:01: UCAST, ttvn 1, ttl 50, IP 10.10.10.2 > 10.10.10.1: ICMP echo reply, id 1588, seq 34, length 64 23:55:18.024977 BAT 08:00:27:00:00:02: OGM IV via neigh 08:00:27:00:00:02, seq 288776459, tq 255, ttvn 1, ttcrc 13868, ttl 50, v 14, flags [...F.], length 26 23:55:18.128469 BAT 08:00:27:00:00:02: OGM IV via neigh 08:00:27:00:00:01, seq 288776459, tq 225, ttvn 1, ttcrc 13868, ttl 49, v 14, flags [.D...], length 46 23:55:18.461849 BAT 08:00:27:00:00:01: OGM IV via neigh 08:00:27:00:00:01, seq 2362693815, tq 255, ttvn 1, ttcrc 52668, ttl 50, v 14, flags [...F.], length 46 23:55:18.564340 BAT 08:00:27:00:00:01: OGM IV via neigh 08:00:27:00:00:02, seq 2362693815, tq 225, ttvn 1, ttcrc 52668, ttl 49, v 14, flags [.D...], length 26 23:55:18.981405 BAT 08:00:27:00:00:01 > 08:00:27:00:00:02: UCAST, ttvn 1, ttl 50, IP 10.10.10.1 > 10.10.10.2: ICMP echo request, id 1588, seq 35, length 64 23:55:18.981457 BAT 08:00:27:00:00:02 > 08:00:27:00:00:01: UCAST, ttvn 1, ttl 50, IP 10.10.10.2 > 10.10.10.1: ICMP echo reply, id 1588, seq 35, length 64 23:55:19.036161 BAT 08:00:27:00:00:02: OGM IV via neigh 08:00:27:00:00:02, seq 288776460, tq 255, ttvn 1, ttcrc 13868, ttl 50, v 14, flags [...F.], length 26 23:55:19.138587 BAT 08:00:27:00:00:02: OGM IV via neigh 08:00:27:00:00:01, seq 288776460, tq 225, ttvn 1, ttcrc 13868, ttl 49, v 14, flags [.D...], length 46 23:55:19.465249 BAT 08:00:27:00:00:01: OGM IV via neigh 08:00:27:00:00:01, seq 2362693816, tq 255, ttvn 1, ttcrc 52668, ttl 50, v 14, flags [...F.], length 46 23:55:19.564327 BAT 08:00:27:00:00:01: OGM IV via neigh 08:00:27:00:00:02, seq 2362693816, tq 225, ttvn 1, ttcrc 52668, ttl 49, v 14, flags [.D...], length 26 23:55:19.987974 BAT 08:00:27:00:00:01 > 08:00:27:00:00:02: UCAST, ttvn 1, ttl 50, IP 10.10.10.1 > 10.10.10.2: ICMP echo request, id 1588, seq 36, length 64 23:55:19.988016 BAT 08:00:27:00:00:02 > 08:00:27:00:00:01: UCAST, ttvn 1, ttl 50, IP 10.10.10.2 > 10.10.10.1: ICMP echo reply, id 1588, seq 36, length 64 23:55:20.024850 BAT 08:00:27:00:00:02: OGM IV via neigh 08:00:27:00:00:02, seq 288776461, tq 255, ttvn 1, ttcrc 13868, ttl 50, v 14, flags [...F.], length 26 23:55:20.128600 BAT 08:00:27:00:00:02: OGM IV via neigh 08:00:27:00:00:01, seq 288776461, tq 225, ttvn 1, ttcrc 13868, ttl 49, v 14, flags [.D...], length 46 ^C
But there are many packets on bat0 which shouldn't be there:
node2 ~ # batctl td bat0 23:55:35.531057 BAT 08:00:27:00:00:01: OGM IV via neigh 08:00:27:00:00:01, seq 2362693832, tq 255, ttvn 1, ttcrc 52668, ttl 50, v 14, flags [...F.], length 46 23:55:36.090997 BAT 08:00:27:00:00:01 > 08:00:27:00:00:02: UCAST, ttvn 1, ttl 50, IP 10.10.10.1 > 10.10.10.2: ICMP echo request, id 1588, seq 52, length 64 23:55:36.091051 IP 10.10.10.1 > 10.10.10.2: ICMP echo request, id 1588, seq 52, length 64 23:55:36.091062 IP 10.10.10.2 > 10.10.10.1: ICMP echo reply, id 1588, seq 52, length 64 23:55:36.192880 BAT 08:00:27:00:00:02: OGM IV via neigh 08:00:27:00:00:01, seq 288776477, tq 225, ttvn 1, ttcrc 13868, ttl 49, v 14, flags [.D...], length 46 23:55:36.529641 BAT 08:00:27:00:00:01: OGM IV via neigh 08:00:27:00:00:01, seq 2362693833, tq 255, ttvn 1, ttcrc 52668, ttl 50, v 14, flags [...F.], length 46 23:55:37.097030 BAT 08:00:27:00:00:01 > 08:00:27:00:00:02: UCAST, ttvn 1, ttl 50, IP 10.10.10.1 > 10.10.10.2: ICMP echo request, id 1588, seq 53, length 64 23:55:37.097071 IP 10.10.10.1 > 10.10.10.2: ICMP echo request, id 1588, seq 53, length 64 23:55:37.097081 IP 10.10.10.2 > 10.10.10.1: ICMP echo reply, id 1588, seq 53, length 64 23:55:37.158529 BAT 08:00:27:00:00:02: OGM IV via neigh 08:00:27:00:00:01, seq 288776478, tq 225, ttvn 1, ttcrc 13868, ttl 49, v 14, flags [.D...], length 46 23:55:37.536005 BAT 08:00:27:00:00:01: OGM IV via neigh 08:00:27:00:00:01, seq 2362693834, tq 255, ttvn 1, ttcrc 52668, ttl 50, v 14, flags [...F.], length 46 23:55:38.103986 BAT 08:00:27:00:00:01 > 08:00:27:00:00:02: UCAST, ttvn 1, ttl 50, IP 10.10.10.1 > 10.10.10.2: ICMP echo request, id 1588, seq 54, length 64 23:55:38.104027 IP 10.10.10.1 > 10.10.10.2: ICMP echo request, id 1588, seq 54, length 64 23:55:38.104037 IP 10.10.10.2 > 10.10.10.1: ICMP echo reply, id 1588, seq 54, length 64 23:55:38.177199 BAT 08:00:27:00:00:02: OGM IV via neigh 08:00:27:00:00:01, seq 288776479, tq 225, ttvn 1, ttcrc 13868, ttl 49, v 14, flags [.D...], length 46 23:55:38.554654 BAT 08:00:27:00:00:01: OGM IV via neigh 08:00:27:00:00:01, seq 2362693835, tq 255, ttvn 1, ttcrc 52668, ttl 50, v 14, flags [...F.], length 46 ^C
Batman 2013.1.0 and Debian kernel 2.6.32-48squeeze3 doesn't show this problem.
Batman 2013.2.0 and Debian kernel 3.2.46-1~bpo60+1 also doesn't show this problem.
These are the attached files
node2 ~ # tcpdump -i eth1 -s 1500 -w node2-eth1.pcap node2 ~ # tcpdump -i bat0 -s 1500 -w node2-bat0.pcap node2 ~ # batctl ll all node2 ~ # batctl l > batman-node2.log
Files
Updated by Antonio Quartulli almost 11 years ago
Hello Oliver,
after a quick reading of your report I have the feeling that there is some compatibility issue between batman-adv and the kernel you are running.
Oliver Möschen wrote:
...
Batman 2013.1.0 and Debian kernel 2.6.32-48squeeze3 doesn't show this problem.
...
Does this mean that batman-adv 2013.1.0 is running fine in the same environment where batman-adv 2013.2.0 is having troubles?
Or the kernel is not exactly the same?
Cheers,
Updated by Oliver Möschen almost 11 years ago
Hello Antonio,
in the same environment batman-adv 2013.1.0 is running fine.
Oliver
Updated by Antonio Quartulli almost 11 years ago
Oliver, do you think you can test 2.6.32 vanilla? In this way we can understand if any of the patches brought by debian is affecting batman-adv somehow.
Updated by Oliver Möschen almost 11 years ago
OK, I will test 2.6.32 vanilla tonight.
Updated by Oliver Möschen almost 11 years ago
I've tested several vanilla kernels with batman-adv 2013.2.0 using the setup described above. These are the results:
2.6.32 Mesh traffic on bat0 2.6.32.27 Mesh traffic on bat0 2.6.33.7 Mesh traffic on bat0 2.6.34.7 Mesh traffic on bat0 2.6.35.9 Mesh traffic on bat0 2.6.36.4 Mesh traffic on bat0 2.6.37.6 Mesh traffic on bat0 2.6.38.8 Mesh traffic on bat0 2.6.39 OK 2.6.39.4 OK
Updated by Antonio Quartulli almost 11 years ago
Thanks for you precious information Oliver.
I will try to understand what's wrong with those kernels. We have a compat layer which makes batman-adv compatible with old kernels despite the API changes brought by the new versions and I think something went wrong there.
We will try to provide you a patch that you can test.
Updated by Antonio Quartulli over 10 years ago
It looks like the problem is in the new RTNL API implementation we introduced in v2013.2.0, in particular in its "compatibility" code which is used when batman-adv is compiled on older kernels.
We should now try to understand what is different in kernels <2.6.39 for what concerns this API.
Updated by Marek Lindner about 9 years ago
- Status changed from New to Closed
I just pushed a patch into the master branch addressing this compat issue. The mailing list has the entire discussion leading up to this patch: https://lists.open-mesh.org/mailman3/hyperkitty/list/b.a.t.m.a.n@lists.open-mesh.org/message/WC6WIL6VTC7OVCY5F7U6KN6DOQTRCSZ5/
Feel free to re-open the ticket if your problem persists.
Actions