Securing the BGP or controlling it?

[vbono at 2nplus1.com (Vincent J.. Bono)]

> this is a matter of risk analysis.  No secure routing means we'll continue
> to see the occasional high profile outage which is dealt with very quickly.

Speaking from painful experience all kinds of variable can ensure that even when a problem is identified quickly and action taken expeditiously outages can and do take much longer than "very quickly" to correct.

Also, while (IMHO) the much higher level of private interconnects / peering links in use today vs. 1997 makes willful route hijacking more difficult, building better security directly into the protocol is certainly in order.  A good parallel is the SS7 network that runs "routing" for traditional voice signaling: it's "secured" by using a completely separate, out of band TDM network (DS1s and DS0s) but its also an "in the clear" protocol and could be subject to willful vandalism.