Vigneswaran R | 21 Jul 14:25 2014

Local attached network info not getting populated

Hello,

In a simple (two nodes) olsrv2 topology, I am advertising gateway for 
0.0.0.0/0 from one node. However, the corresponding route is not getting 
reflected in the neighbor.

This problem was not there with v0.5.0 commit. So, I did a git bisect 
and found that the following commit introduced the problem. Could you 
please help in resolving it?

dda9ae431332fd51726cd373eb1f9664370e062d is the first bad commit
commit dda9ae431332fd51726cd373eb1f9664370e062d
Author: Henning Rogge <henning.rogge <at> fkie.fraunhofer.de>
Date:   Wed Jul 2 08:13:20 2014 +0200

     Small fixes for OLSRv2 parsing/generation

Regards,
Vignesh

--

-- 
Olsr-dev mailing list
Olsr-dev <at> lists.olsr.org
https://lists.olsr.org/mailman/listinfo/olsr-dev

Henning Rogge | 15 Jul 13:23 2014
Picon

Re: olsrv2: observed routing loop when link goes down

On Tue, Jul 15, 2014 at 11:46 AM, Vigneswaran R <vignesh <at> atc.tcs.com> wrote:
> On 07/14/2014 06:39 PM, Henning Rogge wrote:
>> I just pushed several small fixes for NHDP and OLSRv2... you might
>> want to check them.
>
> Yes, it solved the problem :-). Thank you.

After quite a bit of testing and some patches the problem was resolved...

thank you to Vigneswaran for the feedback and testing.

The problem was happening when an interface was going down. This
removed all existing links of this interface directly, which in turn
prevented the node from announcing them as "lost" is the next
Hellos... which made the neighbors NOT throwing their 2-hop database
entries away.

Henning Rogge

--

-- 
Olsr-dev mailing list
Olsr-dev <at> lists.olsr.org
https://lists.olsr.org/mailman/listinfo/olsr-dev

Vigneswaran R | 11 Jul 13:21 2014

Re: olsrv2: observed routing loop when link goes down

On 07/11/2014 04:27 PM, Henning Rogge wrote:
> On Fri, Jul 11, 2014 at 12:51 PM, Vigneswaran R <vignesh <at> atc.tcs.com> wrote:
>> Originator IPs:
>> ---------------
>> vm1 - 10.1.1.1
>> vm2 - 10.1.2.1
>> vm3 - 10.1.3.1
> What are the linklocal IPs and mac addresses on the nodes vm1-vm6?

vm1:
====
eth0: MAC 00:ab:28:2f:87:00, IPv4 169.254.202.130/16, IPv6 
fe80::2ab:28ff:fe2f:8700/64
eth1: MAC 00:ab:28:2f:87:01, IPv4 169.254.7.252/16,   IPv6 
fe80::2ab:28ff:fe2f:8701/64
eth2: MAC 00:ab:28:2f:87:02, IPv4 169.254.224.75/16,  IPv6 
fe80::2ab:28ff:fe2f:8702/64
eth3: MAC 00:00:ab:9c:56:03, IPv4 169.254.65.47/16,   IPv6 
fe80::200:abff:fe9c:5603/64

vm2:
====
eth0: MAC 00:ab:a7:44:02:00, IPv4 169.254.66.163/16,  IPv6 
fe80::2ab:a7ff:fe44:200/64
eth1: MAC 00:ab:a7:44:02:01, IPv4 169.254.17.39/16,   IPv6 
fe80::2ab:a7ff:fe44:201/64
eth2: MAC 00:00:ab:fe:fc:02, IPv4 169.254.167.188/16, IPv6 
fe80::200:abff:fefe:fc02/64
eth3: MAC 00:00:ab:75:13:03, IPv4 169.254.188.126/16, IPv6 
fe80::200:abff:fe75:1303/64
(Continue reading)

Vigneswaran R | 11 Jul 12:21 2014

olsrv2: observed routing loop when link goes down

Hello,

I am running latest olsrv2 in the following topology.

vm1----------vm2---------vm3
| \                      |
|  \______               |
|         \              |
|          \             |
vm4----------vm5---------vm6

Fig.1  Six qemu virtual machines having four Ethernet NICs each

* Initially, route from vm1 to vm6 is via vm5 ("vm1-->vm5-->vm6")
* I started ping from vm1 to vm6
* While ping was in progress, I put vm5's interface which is connected 
to vm6, DOWN (just to simulate 'vm5-vm6' link break)

At that point, ping gave a couple of "Destination Net Unreachable" 
error. After that it started giving "Time to live exceeded". This 
persists until I make 'vm5-vm6' link up.

During that period, vm1 had route to vm6 via vm5, and vm5 had route to 
vm6 via vm1. That created a loop.

Note:
----
If I drop the packets using iptables to simulate the link break (instead 
of putting the interface down), olsrv2 is able to recover from the 
routing loop quickly (in 4 or 5 secs).
(Continue reading)

Patrick Grimm | 10 Jul 18:41 2014
Picon

[PATCH 1/2] fix jsoninfo too listen on ipv6 port only

If not IPV6_V6ONLY flag ist set then olsrd with ipversion=6
is listen on IPv6 and IPv4 tcp port. A second process with
ipversion=4 can’t listen on IPv4 tcp Port.

Signed-off-by: Patrick Grimm <patrick <at> lunatiki.de>
---
 lib/jsoninfo/src/olsrd_jsoninfo.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/lib/jsoninfo/src/olsrd_jsoninfo.c b/lib/jsoninfo/src/olsrd_jsoninfo.c
index d86fa88..0db7609 100644
--- a/lib/jsoninfo/src/olsrd_jsoninfo.c
+++ b/lib/jsoninfo/src/olsrd_jsoninfo.c
 <at>  <at>  -385,6 +385,14  <at>  <at>  plugin_ipc_init(void)
       return 0;
     }
 #endif /* (defined __FreeBSD__ || defined __FreeBSD_kernel__) && defined SO_NOSIGPIPE */
+#if defined linux
+    if (olsr_cnf->ip_version == AF_INET6) {
+      if (setsockopt(ipc_socket, IPPROTO_IPV6, IPV6_V6ONLY, (char *)&yes, sizeof(yes)) < 0) {
+        perror("IPV6_V6ONLY failed");
+        return 0;
+      }
+    }
+#endif /* defined linux */
     /* Bind the socket */

     /* complete the socket structure */
--

-- 
2.0.1
(Continue reading)

Patrick | 10 Jul 00:25 2014
Picon

[PATCH] olsrd Fix listen on ip6 only in jsoninfo


Hi

I cant start 2 process of olsrd with ip6 and ip4.
The ip6 process is listen on ip4 tcp port 9090.

I'm not sure if this the right places to do this.

diff --git a/lib/jsoninfo/src/olsrd_jsoninfo.c b/lib/jsoninfo/src/olsrd_jsoninfo.c
index d86fa88..0db7609 100644
--- a/lib/jsoninfo/src/olsrd_jsoninfo.c
+++ b/lib/jsoninfo/src/olsrd_jsoninfo.c
 <at>  <at>  -385,6 +385,14  <at>  <at>  plugin_ipc_init(void)
       return 0;
     }
 #endif /* (defined __FreeBSD__ || defined __FreeBSD_kernel__) && defined SO_NOSIGPIPE */
+#if defined linux
+    if (olsr_cnf->ip_version == AF_INET6) {
+      if (setsockopt(ipc_socket, IPPROTO_IPV6, IPV6_V6ONLY, (char *)&yes, sizeof(yes)) < 0) {
+        perror("IPV6_V6ONLY failed");
+        return 0;
+      }
+    }
+#endif /* defined linux */
     /* Bind the socket */

     /* complete the socket structure */
Henning Rogge | 9 Jul 14:12 2014
Picon

Re: olsrv2 routes with same source IP regardless of outgoing interface

On Wed, Jul 9, 2014 at 2:09 PM, Vigneswaran R
<r_vigneswaran <at> users.sourceforge.net> wrote:
> Yes. Now olsrv2 is not assigning any IPv4 address automatically, if there is
> no IPv6 address present on the interfaces. Thanks.

Thank you for all the testing... without your input I wouldn't have
found a few bugs I was able to fix.

Henning Rogge

--

-- 
Olsr-dev mailing list
Olsr-dev <at> lists.olsr.org
https://lists.olsr.org/mailman/listinfo/olsr-dev

Henning Rogge | 9 Jul 12:19 2014
Picon

Re: olsrv2 routes with same source IP regardless of outgoing interface

On Tue, Jul 8, 2014 at 10:41 AM, Vigneswaran R
<r_vigneswaran <at> users.sourceforge.net> wrote:
> May be you can flush the IPv6 addresses ('ip addr flush $iface') for
> testing. Because, that's what happened in my case -- I flushed both IPv4 and
> IPv6 addresses without realizing that IPv6 addresses are used for auto
> configuration of IPv4 address.

Okay,

I did some bug hunting and fixing and I think it works for me with the
most recent commit.

Henning Rogge

--

-- 
Olsr-dev mailing list
Olsr-dev <at> lists.olsr.org
https://lists.olsr.org/mailman/listinfo/olsr-dev

Henning Rogge | 8 Jul 09:19 2014
Picon

Re: olsrv2 routes with same source IP regardless of outgoing interface

On Tue, Jul 8, 2014 at 9:11 AM, Vigneswaran R
<r_vigneswaran <at> users.sourceforge.net> wrote:
> On 07/08/2014 12:13 PM, Henning Rogge wrote:
>>
>> Hi,
>>
>> could you test if the attached patch resolves the problem on "ipv4
>> only" interfaces?
>
>
> Hmm.. didn't get success..
>
> I explicitly flushed the ip addresses (both IPv4 & IPv6) from the interfaces
> and started the patched olsrd2 daemon. Still it is assigning the same IP
> addresses 169.254.1.0/16 on all the interfaces of the virtual machine
> (tested with one machine only).
>
> The only difference is, previously when I kill olsrd2, it removes the IP
> addresses it added. Now, the IPs are not being removed. PFA the olsrd2 log
> file.

Okay, this seems to be a more complex problem than I thought, I have
to setup a few IPv4 only virtual machines.

I will contact you again if I have a better solution.

Henning Rogge

--

-- 
Olsr-dev mailing list
(Continue reading)

Vigneswaran R | 25 Jun 12:35 2014

olsrv2 routes with same source IP regardless of outgoing interface

Hello,

I observed that the routes added by OLSRv2 are having the originator IP 
address as the source IP, regardless of the outgoing interface. (With 
olsrv1, the IP on the outgoing Interface is used as the SRC IP). Sample 
OLSRv2 routes,

10.1.2.1 via 10.1.2.1 dev eth0  proto 100  src 10.1.1.1  metric 2 onlink
10.1.4.1 via 10.1.4.1 dev eth1  proto 100  src 10.1.1.1  metric 2 onlink
10.1.5.1 via 10.1.5.1 dev eth2  proto 100  src 10.1.1.1  metric 2 onlink

The following lines (src/olsrv2/olsrv2_routing.c) seem to confirm that 
as the expected behaviour. Is that correct?

     if (_domain_parameter[rtentry->domain->index].use_srcip_in_routes
         && netaddr_get_address_family(&rtentry->route.dst) == AF_INET) {
       memcpy(&rtentry->route.src, olsrv2_originator_get(AF_INET),
           sizeof(rtentry->route.src));
     }

Regards,
Vignesh

--

-- 
Olsr-dev mailing list
Olsr-dev <at> lists.olsr.org
https://lists.olsr.org/mailman/listinfo/olsr-dev

Ufo | 20 Jun 16:45 2014
Picon

http-info down since some days (luci)

Since some days our trunk-Firmware doesnt provide the Display of OLSR 
Neighbors via Webinterface any more. :-( (luci, with and without root-login)

"unable to connect to the olsr daemon!"

plugin is active, olsr via ipv6 is still disabled, but the 
commandline-script via "neigh.sh" is still working.
Does anyone have same issues? I suppose its since these troubles when 
changing/renaming config-files for olsr.

p.s. does anyone has started programming a webinfo about olsr-neighbors 
within luci2? we would like to test it to get rid of luci1.

greetings, ufo

--

-- 
Olsr-dev mailing list
Olsr-dev <at> lists.olsr.org
https://lists.olsr.org/mailman/listinfo/olsr-dev


Gmane