GoDaddy: We Were Not Attacked, It Was a Server Problem

[blazorthon]

[citation][nom]A Bad Day[/nom]He has better English than some of the people around here sadly.[/citation]

To be fair, this sentence of yours isn't proper English either.

 

[house70]

[citation][nom]blazorthon[/nom]To be fair, this sentence of your isn't proper English either.[/citation]
To be fair, neither is yours....

 

[blazorthon]

[citation][nom]house70[/nom]To be fair, neither is yours....[/citation]

Point made and accepted; I had a typo too. I'll fix it.

 

[Guest]

I can believe that it was a routing mistake. I've been studying for my CCNA and I can see how easy it is to mess up your core network routers with a mistaken setting. Cisco routers, by default, assume a static route you manually enter in is superior to any dynamic routes learned from other routers. So one bad static route and traffic that was supposed to go to their DNS servers to get the host addresses ends up going nowhere.

If it was a server problem, there's lots of tools to undo or roll-back a change, or even restore from a backup. Nothing to that degree exists to prevent a newb from trashing a router's settings.

 

[ozvip3r]

[citation][nom]blazorthon[/nom]That "is" and the "and" shouldn't be there. The "and" also gives the impression that there's a word missing rather than an extra "and".On topic:Any network admins out there, can you describe what would need to be done to hack a system and do something like this? It seems like if it was an attack, it was more complex than the average script kiddie can manage and the response on twitter doesn't seem how someone who's better than a script kiddie would word a response.EDIT: added another mistake that was made in the article.[/citation]
These blatant errors are on every 2nd article if not more. They are so visible, such as this one at the very beginning. They don't even bother to proof read or hire competent writers with basic grammar skills, like sentence structure 😛 It has become a joke over the last year or so!

 

[A Bad Day]

[citation][nom]blazorthon[/nom]To be fair, this sentence of yours isn't proper English either.[/citation]

Go away grammar police...

 

[blazorthon]

[citation][nom]A Bad Day[/nom]Go away grammar police...[/citation]

You're the one who was commenting on the grammar from some of the "people around here", I simply commented on your comment of it.

 

[JamesSneed]

I usually don't care but it is pretty bad when the title says server problem and the article says router problem. Take that along with the very first sentence having two very annoying mistakes and I think its worth noting. Toms, please try to do better.

 

[rantoc]

Or that "internal network events" could have been triggered from the outside, if the it staff is indeed so "professional" that they can cause a fault of this magnitude and it even took them forever to fix it... i doubt their ability to find the cause of the problem in the first place. Sure its easier to blame whatever internal than say we got hacked.

I don't trust either party here, it could have been annon who triggered the fault or it could be go daddy who take the less pr unfriendly way of trying to smooth any such events out.

 

[acadia11]

[citation][nom]hapkido[/nom]According to godaddy, they messed up their routing tables. It wouldn't be an issue of overloading servers. It's like this --If router A is directed router BAnd router B is directed to router CTraffic will flow from A to CNow let's same the entry became corrupted or (more likely) someone made an error.If router A is directed to router BAnd router B is directed to router ATraffic will not flow from A to CThat's a simplified view, but that's what they're saying happened, and it's very plausible that's what did happen. It's more likely that a careless sys admin committed a bad routing entry than anonymous DDoSed godaddy.[/citation]

Well then they must have committed it on quite a few routers. 1 router isn't going to control let's just guestimate 5 million sites / 1000 = 5000 nodes (servers), so ... my point 1 router isn't going to be the router for 5000 servers. While it's plausible they have some mechanism that updates all the router tables, it would have to be quite a few routers that were improperly configured to take down that large of the network.

 

[freggo]

"We have determined the service outage was due to a series of internal network events that corrupted router data tables"

Did GoDaddy clear this with their lawyers?
By admitting it was their fault they open themselves up to lawsuits for damages of course !