They could have dragged twenty or thirty web servers out of inventory, racked them, mirrored an image from a production machine to the new ones, modified the configs, and added them to the load balancer.
Originally Posted by porkozone
If they were already clustering the DB, throw two more servers at it and update the cluster. If they weren't, either rush a cluster, or rush a replacement of the current storage with something state-of-the-art out of inventory.
HP is one of the most popular manufacturers of web serving equipment in the world-- they had the stock sititing on a loading dock somewhere, and they would have been better off ****ing off one guy trying to run some start-up than ****ing off thousands of customers.
I worked for one of the main movie ticketing sites-- if we could sell 20K tickets/hour back in 2004, HP could have managed to move 100K TPs without crashing over a weekend.
But they failed to plan for it, then they failed to react to it.