What happens when you outgrow existing infrastructure? You hit a wall and it slows you down. I can imagine a team of tech gurus hovering over server delay, outage and DNS reports — half the team insisting they knew they would hit this wall eventually, the other half just trying to get the problem solved. So what is the problem? Is it just not enough servers, processing power and load-balancing to keep up with the Facebook demand? Obviously. But where did the failing point happen? Within the company server rooms or in the connection to the back-bone to the Internet? Curious how many co-location sites these guys have, where they might be?
Really, I only noticed the site being down a couple of times today but what I could not ignore was all the hype and fuss about what was going wrong. Interesting. Ah, back to the real world. =)