mSQL Attack/Peering/OBGP/Optical exchange

Of course, I realize that to implement the necessary rules would add a
complexity that could cost largs sums of money due to mistakes.

Implementing the automation that can (correctly) implement the
necessary rules is an enormous challenge, and it's unclear whether
anyone in the marketplace will rise to meet it; the market for those
additional (hard to implement) features is small, while the less
complex features that just implement the auto-provisioning
spend-o-rama between a vendor and its customers in the aforementioned
hierarchical manner are easier to implement and have a broader market.
Companies trying to make a profit in the software development game are
naturally repelled by the former and attracted by the latter,
especially when their target is a multi-vendor environment. Vendors
making software to manage their gear and their gear alone often
concentrate more on ringing all the bells and tooting all the whistles
of their device rather than implementing a plethora of policies
(sorry, I had to say it) cleanly and simply.

Custom code can fill in the gaps, as we've seen so often in the
service provider market. Once in a while, something truly useful
escapes into the wild, like rancid, but the really good stuff tends to
stay safely eithin the walls of the organization that developed it
because it represents a competitive advantage.

VP, Eng.