, , , ,

The obvious news from the last 48 hours was the death of Michael Jackson. The first thought that I gave was the obvious battering that the likes of Ticketmaster and Seatwave were going to get. Initially it was the purchase of tickets, now it’s for the refunds.

Any large scale event will alway incur a spike at the time tickets go on sale. Something like:


(Apologies for my crude Google Chart example).

It stands to reason that an elastic system would be of benefit to any high volume ticketing system. Deploy n number of servers to cope with the demand. Michael Jackson’s O2 shows were a good case in point. There in the region of 250,000 tickets on sale and they went in a matter of hours. If those 250,000 sales represented 1% of the hits the site was getting then you can easily estimate that the server could potentially get 25,000,000 hits in a short space of time.

There’s also a maximum return on investment (ROI) from only having the servers you need available at peak times, instances then switched off during quieter periods. Saves having a server farm full of running servers.