On Mon, Aug 18, 2014 at 09:21:30pm -0700, Christian Huitema wrote:
> >> That means preventing RST injection.
> >
> > See above for the consequences of this.
> 
> I think that by now, we all understand the tradeoff. If we want to prevent 
> illegitimate RST attacks, we will also prevent use of unprotected RST by 
> rebooting hosts. Which is why I believe the protocol should be able to 
> differentiate between protected RST, which should be processed immediately, 
> and unprotected RST, which should be treated according to host policy.

Yeah - that does strike me as about the best we can achieve for RST.

Maybe some recommendations for the policy,  but ultimately down to policy for 
the unprotectable RST cases.

.pdf

_______________________________________________
Tcpinc mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/tcpinc

Reply via email to