[Olsr-users] Where am I going wrong in the settings.

[Teco Boot]

OK, your dyn_gw works just fine.
The interface RUNNING status is described here: https://www.kernel.org/doc/Documentation/networking/operstates.txt
If you need to supported plugged and unplugged ethernet ports, I recommend ifplugd. I use it for managing dhcp client.

Teco

Op 11 apr. 2014, om 14:13 heeft ANDERSON JUNIOR GADO DA SILVA <(spam-protected)> het volgende geschreven:

> hello, this olsr date with svn, latest version.
> 
> right, follow the steps you gave me and then post the results
> 
> I did everything exactly as you said, to take out the ping fails, in the early hna the 2nd ceases to be announced
> 
> echo "/ hna" | nc 127.0.0.1 2006
> HTTP/1.0 200 OK
> Content-type: text / plain
> 
> Table: HNA
> Destination     Gateway
> 
> the first time I waited 2 minutes and the second test I waited 10 minutes, the default route with metric 0 172.16.0.1 continued it, but now as a node only had then just got 172.16.0.1, to reconnect the cable in a few seconds the ping back working and hna shall be reposted without changing anything
> 
> echo "/ hna" | nc 127.0.0.1 2006
> HTTP/1.0 200 OK
> Content-type: text / plain
> 
> Table: HNA
> Destination     Gateway
> 0.0.0.0 / 0       10.5.0.5
> 
> the problem is that the default route does not leave the table, even when she's not working anymore.
> 
> sorry about the time I have taken from you. I am available for any test needed
> 
> Att
> Anderson Junior
> 
> 
> 
> 
> 2014-04-11 3:29 GMT-04:00 Teco Boot <(spam-protected)>:
> 
> Op 11 apr. 2014, om 08:33 heeft Henning Rogge <(spam-protected)> het volgende geschreven:
> >
> >> 2. dyn_gw with ping check doesn’t work.
> >> Long ago I helped Caspar to repair this plugin. I tested, it worked. Let’s see what we can do.
> >> I tested, below my results. It works.
> >> Please verify with same configuration and show your results.
> >> Do you have tcpdump?
> >
> > *looking at next mail*
> >
> > the tcpdump looks fine, ICMPs to Google are not working anymore when the
> > cable is pulled and the HNA is dropped.
> >
> > It would take some time for the HNA entries in the other nodes to
> > timeout, but after ~30 seconds they should be gone.
> 
> Yes, it would take some time.
> This was just the reference test dyn_gw is working.
> 
> 
> @Anderson:
> Could you test with single node, with ping test to 8.8.8.8, in order below. Run tcpdump in background.
> 1. olsrd is running, cable connected, default route is in routing table, ping to 8.8.8.8 succeeds, HNA is advertised
> 2. remove cable, ping to 8.8.8.8 fails
> 3. each 10 seconds, check routing table for default routes (metric 0 and metric 2), HNA advertisement
> 4. after 2 minutes or so, connect the cable and reinstall your static route if needed, ping to 8.8.8.8 should succeed
> 5. repeat step 3.
> 
> I did see output, where OLSR reacted on disconnected cable. But there was also the static route. This was strange, please confirm if this problem is still present.
> 
> Teco
> 
> 
>