According to an article on TechCrunch, GoDaddy’s response is:
We have determined the service outage was due to a series of internal network events that corrupted router data tables.
I’m not certain that’s a more comforting response than, “we got h4ck3d!” Some people, like reddit user snappywan, are calling BS.
To quote snappywan:
From a network admin point of view, their response is bullshit. They just don’t want to admit there may or may not have been a security breach which would stir sensationalist headlines in the media, scaring their entire customer base, and result in a larger loss of customers. To be fair, this is probably the safest business practice they could have done.
See, they were reachable by IP during their DNS outage. So the routing tables were perfectly fine, and their routers were completely operational. Proof: use  BGPlay to view the history of route announcements on the Internet, select sep 10 from 00:00 to 23:00, for 126.96.36.199/22 (one of godaddy IP range)). As you will, there were no route withdrawals, only reannouncements.
I can’t help but think it’s weird when I see a client spend thousands of dollars for a website, but their “web developer” chose to use GoDaddy for web hosting. It’s difficult to argue with someone who defends using GoDaddy for web hosting, DNS, email, etc.
Sure, sites go down and ALL service providers experience some form of issue at some point. 99% up-time still means there can be some down-time! For those who have been supporting clients for years though, they know the issues brought about by using GoDaddy’s web hosting and email services – e.g. email delivery issues, blank emails, “obfuscated” emails, performance issues, etc. Don’t get me started on trying to navigate through GoDaddy’s checkout process.
For the geeks playing along at home, let’s not forget to mention RFCs.
It’s one thing to register a domain there, but to dump all of your eggs in one basket, is not a sound approach.