In addition to searching for Web pages, Google users can search for PDF, PostScript, text, Microsoft Office, Lotus, PowerPoint, and Shockwave files. Google claims to be one of the five most popular sites on the Internet with more than 380 million unique users per month and more than 50 percent of its traffic coming from outside the United States. Google’s IT infrastructure is a closely guarded secret because it is part of its competitive advantage. The best guess is that Google has up to 450,000 servers spread over at least 25 locations around the world. These servers use inexpensive off-the-shelf hardware to run a customized version of the Linux operating system and other critical pieces of custom software. These include MapReduce, a programming model to simplify processing and create large data sets; Google WorkQueue, a system that groups queries and schedules them for distributed processing; and the Google File System, which keeps copies of data in several places so that the data will always be available even if a server fails. According to a widely cited estimate, Google only needs to spend $1 for every $3 its competitors spend to deliver a comparable
In addition to searching for Web pages, Google users can search for PDF, PostScript, text, Microsoft Office, Lotus, PowerPoint, and Shockwave files. Google claims to be one of the five most popular sites on the Internet with more than 380 million unique users per month and more than 50 percent of its traffic coming from outside the United States. Google’s IT infrastructure is a closely guarded secret because it is part of its competitive advantage. The best guess is that Google has up to 450,000 servers spread over at least 25 locations around the world. These servers use inexpensive off-the-shelf hardware to run a customized version of the Linux operating system and other critical pieces of custom software. These include MapReduce, a programming model to simplify processing and create large data sets; Google WorkQueue, a system that groups queries and schedules them for distributed processing; and the Google File System, which keeps copies of data in several places so that the data will always be available even if a server fails. According to a widely cited estimate, Google only needs to spend $1 for every $3 its competitors spend to deliver a comparable