Update
ClickFrenzy finally managed to get a page up. Nice work on trying to market the downtime as a positive thing when really ClickFrenzy a national joke at this point.
Tonight should have been Australia’s biggest online sales event in history. Marketed across the web and mainstream media, ClickFrenzy was supposed to provide big discounts from a large number of Aussie online providers. The problem is that at 7PM when the site went live, the traffic influx was too much for the server to handle.
Essentially this was an invited DDOS attack. While ClickFrenzy selected UltraServe to serve up their site, it seems their quoted ‘Ahead in the cloud’ tag line isn’t exactly accurate. Given the expected influx of traffic at a specific time (we’re talking potentially millions of simultaneous connections), a serious large scale provider should have been selected.
On offer is Microsoft’s Azure platform, Amazon’s EC2, or even Australia’s very own Ninefold distributed cloud hosting. Dramatic spikes in traffic like this event is the perfect use case for the scalable capacity of the cloud. There is one variable in the equation and that is cost.
As the person who’s holding the attached credit card, it would be a very scary bill to be facing the day after ClickFrenzy ends and the traffic subsides. Although it technically would be possible to keep the site online during large influxes of traffic like this, the cost associated would be incredibly expensive.
While there was millions of dollars in transactions at stake here, it’s unclear if ClickFrenzy had negotiated any cost recovery measures from those retailers involved. For now, enjoy a proxy error, or 503, or any of the other array of error messages being reported on twitter using #ClickFrenzy.
Lesson learnt people, this is not how to hold a major online event.
