On Sun, Feb 03, 2002 at 06:02:59PM -0800, Will Lowe wrote: > I understand the "people who have broken routers or block ICMP for > security reasons should get their act together" argument. Beyond > that, what's the argument against it? Why does the above note claim > "it does not work in any case"? Try to formulate an good algorithm for it and you'll see. When do you want to declare a non working connection pmtu holed? There is no non ambigious way to detect a pmtu blackhole vs a network outage. All the possible reactions (turning DF off, shrinking PMTU by guessing) have bad drawbacks given this ambiguity. In practice everybody seems to just rewrite MSS options instead. -Andi - : send the line "unsubscribe linux-net" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
