[Openswan Users] Disappearing %defaultroute

Mario Lobo mario.lobo at ipad.com.br
Thu Oct 20 10:46:52 CEST 2005


Hi;

I know this is in the FAQ. I was just wondering if there is any news on patches, fixups, etc... 

I am using kernel 2.6.13.4 with all klips patches applied. I tried using ipsec.ko but pluto kept 
seg faulting, so i compiled ipsec into the kernel. (no nat-t patches).

using openswan 2.4.0. 

Everything works !!. Even with one end of the tunnel behind 2 NATs. I can ping all machines on the 
LANs at both ends, from both ends. 

LAN1 <--> GATEWAY(NAT) ipsec0 <--> ROUTER(NAT) <--> (internet) <-->  ipsec0 (NAT)GATEWAY <--> LAN2

I left (overnight) a machine on LAN1 pinging a machine on LAN2 and vice-versa. When I arrived this 
morning, both machines had stopped at around 5000 pings. When I checked the routes, the 0.0.0.0 
route was going out through ipsec0 instead of eth0 interface on both GATEWAYs.

I tried:

ipsec tncfg --attach --virtual ipsec0 --physical eth0 (yes, eth0 IS my outter inerface)

the command performmed with no erros, but the default route was still through ipsec0. I tried 
detach and re-attach but, results were the same.

Only a service network restart solved the problem. 

I know it is easy to solve this with a cron script but I would like to avoid that. It would be 
better to find out the whys and whats of this problem, and try to solve it from inside.

So I wrote to see if there is any new info on it.

Thanks,
-- 
   //|  //||
  // | // ||
-//--//---|| ARIO LOBO
//  //    ||
---------------------------------
mario.lobo at ipad.com.br
http://www.ipad.com.br



More information about the Users mailing list