Glastonbury tickets went on sale this morning and I didn't even get through to a holding page despite having multiple tabs open with their own cookies and refreshing like crazy. I got 'This webpage is not available' error in Chrome when each request just timed out.
It got me wondering what determines what connection gets through on a high load. Now I know there's many potential bottlenecks with load balancers and ticket databases on the backend. But say you've got one 1 Apache instance serving static HTML content and it's getting hit with huge amounts of requests that are all timing out. What determines which ones get through and are served the HTML?
I think FIFO (first in first out) applies here, when you connect to a website with an overloaded CPU you will just be waiting in queue for your request to get processed, if it timeout during this time you will have to resend the request. take a look at this http://blog.scoutapp.com/articles/2009/07/31/understanding-load-averages