Thats weird, because I work for a FTSE 100 financial services company and we had a few little issues when nearly 72,000 members of the public attempted to use our online services at the same time, when our usual concurrent volume is around 4,500.
We planned for a "worse-case" of 50,000 concurrent trades, but our system DDOS-ed itself at around 71,000 (gotta love fast-switch cluster failover, and back, and over, and back... Oracle FTW!!!)
Anyhow, before we could even think to phone Oracle and shout that our newest 1.2 million pound server wasn't handling particularly well, the head of Oracle UK phoned /us/ and offered us a team of their best engineers to help get us fixed.
I'd have thought we'd be smaller than a US gov agency, but hey ho, Oracle (plus Cisco, Juniper and EMC) parachuted people straight in, nothing too much to help.
Admittedly now we have bought a lot more kit, but they weren't to know that we'd do that considering the failure.