Re: [opennhrp-devel] Some problem with dead peer

[Timo T. <tim...@ik...>]

Hi,

ChikinAD wrote:
> Can i ask you about dead peer detection?
> 1) Established connect to cisco, all ok
> 2) in the cisco console i clear all sessions (clear cry isa sa) or reload it to simulate connection failture
> 3) racoon reestablishes connections but nhrp is dead, if i restart opennhrp everything is ok!
> How can i do this automaticaly?

You need to make racoon tell opennhrp if the ISAKMP SA dies.
Basically that happens via phase1 script. There's an example script called
racoon-ph1down.sh in opennhrp tarball. And you need to specify it in the
racoon.conf via 
	script "/path/to/racoon-ph1down.sh" phase1_down;
inside remote block.

- Timo