The Braseltons

I was reading an article from Joe Stump on Distributed vs Redundant systems and started geeking around.

I chose Google and how I would implement the their entire model. I read another article where Google serves up 293 Million queries a day (March 2009 numbers). I looked at that number and thought it was unimpressive.

I then began to calculate queries per second. It is 3,391 queries per second!!! Wowsers! They have to return the information in .0003 seconds or less! The dataset is, well, THE INTERNET!

Just for funsies, I ran a search on a MySQL database that wasn’t that large. It returned the results in .1412 seconds. That means that it would take 479 seconds for that computer to handle 3,391 queries. That’s 8 minutes for what Google spits out in a second! Their dataset is GIGANTIC to boot!

I figured a completely redundant method for databases, crawlers and front end servers and how to spit that out across the USA. It’s not all that bad. Just money and time!