Yeah, I know I am weeks late writing about this, and I’ve even written about other things in the mean time…. Well, sorry, but this one just kind of kept getting pushed back, unfortunate ’cause it’s important!
Ticketmaster had a meltdown when they recently put tickets on-sale for the summer 2008 Beijing Olympics. There have been a number of articles about it (ValleyWag.com, News.com, Wall Street Journal). In the end, they had to suspend the on-sale because it appears that the servers couldn’t handle the glut of requests….
Why would they have a meltdown you wonder, in this day and age of huge server farms, etc. why would they not be able to handle the volume…? Well, I guess that it wasn’t the front end, the web service/interface, that had the problem. The Ticketmaster system is a hodgepodge of old and new. While the web interface and all the fancy stuff runs well, runs on the front-end and does so on relatively new technology and servers… the “host system” which is the actual ticketing software (or at least the meat and potatoes of it) is where the bottle neck happened.
Don’t get me wrong, the Ticketmaster host system is an amazing thing for what it does and for it’s age. The bottle neck, if it was even caused at this point, was in no part due to the host handling transactions, i am guessing it was about “user” capacity to get on the host system. The one weakness that I could think of in this type of situation is the host not being able to handle that many “users” concurrently….. ah, legacy systems :-)!
They will get it figured out I’m sure… or if not, they will find a work-around. Until then, I’ll keep writing and telling you my thoughts, until I finally get around to doing ticketing “right” and developing a new system and even more importantly, business model for ticketing :-)!!!