Re: Question about REAP state transition (draft-ietf-shim6-failure-detection-09)

[Iljitsch van Beijnum <iljitsch@muada.com>]

On 14 feb 2008, at 13:34, Alberto García wrote:

> Some time after A is not receiving anymore data from B, the Send  
> timer at A
> expires. Then, A moves from Inbound_ok to the Exploring state,

The way I see it, you don't go from Inbound_OK to Exploring. The text  
"received incoming probe since the last transition from Operational to  
Exploring" is consistent with that and not with the situation where  
Inbound_OK -> Exploring is possible.

However, the advantage of allowing Inbound_OK -> Exploring is that  
this can work around the situation where a second failure breaks even  
more a short time after the first failure.

On the other hand, this should be rare and eventually, the protocol  
will converge on what's still working even though selecting a new  
unidirectional address pair that stops working before the address pair  
for the other direction will lead to some extra delays.

Iljitsch