On Tue, Jun 15, 2004 at 09:26:58PM +0100, Richard Ibbotson wrote:
> Have a read of the logs below and let me know what you think it might 
> be...

There are two (probably unrelated) strange things in there:

> Jun 15 21:09:30 debroglie /netbsd: ippp0: lcp: RXJ+ (proto-rej) for 
> proto 0x8057 (ipv6cp/req-sent)
> Jun 15 21:09:30 debroglie /netbsd: ippp0: ipv6cp close(req-sent)

Your peer did not like to talk IPv6 with you, so ipv6cp is rejected
and closed...

> Jun 15 21:09:30 debroglie /netbsd: ippp0: ipv6cp output <term-req 
> id=0xce len=4>

... but we should not send this term-req then. And we insist on it,
by retrying:

> Jun 15 21:09:31 debroglie /netbsd: ippp0: ipv6cp TO(closing) 
> rst_counter = 2
> Jun 15 21:09:31 debroglie /netbsd: ippp0: ipv6cp output <term-req 
> id=0xcf len=4>

until rst_counter is 0. 

The other thing is, while everything looks fine, business as usual:

> Jun 15 21:10:22 debroglie /netbsd: ippp0: got lcp echo req, sending 
> echo rep
> Jun 15 21:10:22 debroglie /netbsd: ippp0: lcp output <echo-reply 
> id=0x3 len=12 56-f3-2f-43-56-f3-2f-43>

All right, peers probes if we are still here - and we answer. But suddenly:

> Jun 15 21:10:29 debroglie /netbsd: ippp0: lcp down(opened)

and the rest follows from that. There is no indication of idle timeouts
or whatever that caused this "down" event.

Nothing obvious. I suppose you did not do a "ifconfig ippp0 down" at this
moment ;-)

What OS version are you running?

Martin