Yoav Nir writes:
> As a general principle, yes. But the HA extension already assumes
> that due to the failover, there is some discrepancy. The easy way
> out would be to write a protocol extension that just detects this
> discrepancy and kills the IKE SA. But that would mean a lot of IKE
> SA setups following a fail-over. 

I do not think we need such protocol. 

> So in the context of the HA extension, we are willing to live with
> some unsynchronized state, and try to let it take care of itself.

I agree on first step, but I would rather like to see some text
explaining how they are taken care of. I do not think we can assume
the "problem takes care of itself". 

> So the question is, if the peer already does not have the IPsec SA,
> does it add any information for us to send it a DELETE? 

In IKEv2 context it is mandatory, but in high availability IKEv2
context we can specify whatever we want, but we need to specify what
we want. 
-- 
[email protected]
_______________________________________________
IPsec mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/ipsec

Reply via email to