Re: Autonomous System Sanity Protocol

[William Allen Simpson <[e-mail]>]

Unfortunately, connectivity was down all weekend to morningstar (until
monday afternoon), and all these messages have been propagating to my
mailbox all night....  Catching up:

> From: Andrew Partan <[e-mail]>
> >     [The problem came in two parts. ... The second part is that the routes
> >     didn't get widthdrawn correctly - they still existed in various parts of
> >     the Internet some 24 hours after B disconnected itself from the Internet.]
> >
> > Right, and this was the part I found most interesting. Of course, I'd claim
> > that failures like this are almost inevitable in DV type systems! :-)
>
> This was a failure in the propogation of the routing information
> - such failures can happen in DV or MP or any routing system.
>
One of the features of securing the routing information would be to
prevent these software bugs in the first place.  In this case, the bad
routes would not have been believed, and therefore not propagated.  The
bugs would have been detected much sooner (before shipping), as simple
tests would have prevented any routing from occuring.  An ounce of
prevention is worth a pound of cure.

And I agree with Noel, these particular forms of bug (bogus announcement
with truncated path, and failure to withdraw) were both more likely to
occur with DV style propagation.

[e-mail]
    Key fingerprint =  17 40 5E 67 15 6F 31 26  DD 0D B9 9B 6A 15 2C 32
[e-mail]
    Key fingerprint =  2E 07 23 03 C5 62 70 D3  59 B1 4F 5E 1D C2 C1 A2