This blog is somewhat famous for beating up single points of failure and dragging them through the dirt. The fact that anyone would succumb to such a restriction shows either ridiculously poor judgement or the details are buried deep within the 16 pages of legalese nonsense that only lawyers can engage in massively bleeding the company without legal representation of any bank balances.
So what happens when your carrier is an SPF?
Would you be surprised that most of the large networks happen to be SPFs? (the whole SPFism, from the megalomaniac carriers to IT consultants, is a condition borne out of extreme arrogance – “we have the best network, we know it all and we refuse to hire anyone else to transit our data – the whole world should change to adapt to us” which is about as remote from the purpose of the Internet as possible)
How do they get away with it?
Somewhere in the book of terms and conditions of the agreement you sign with a major telecom organization there is text that basically reads:
“Due to the technical limitations and intermittent outages of certain routes ____ does not guarantee transit to all remote networks.”
This happens to be a very valid argument on surface. Suppose there was a web server somewhere in Malaysia and the boat that it’s on just sunk 2 feet and the switch that was at the bottom of it flooded. Of course we can’t guarantee that.
But when Sprint, an SPF network, decides to drop their peering with Cogent, the worlds largest carrier of Internet’s primary value (pr0n, baby) many people start a bond fire.
If you are reading this blog and happen to have a T1 from Sprint you need to drop out of that contract immediately and move to a multihomed carrier. This tactic of bullying your competitors is not just unfair it is a sign of financial instability.
Disclosure: My employer is a large customer of both carriers, however, we do not use either network in an SPF configuration, all are multihomed.