Google uses a process called redundancy to ensure that data will not be lost on its servers. Data on each index server is broken down into 64-megabyte blocks. Three copies of each 64 MB data block are stored on separate servers that run on their own power sources. The data blocks are distributed in a random way so that two servers never have the same collection of blocks. Thus the data will still be available in other machines if there are problems with any particular server.
What is information technology?
Answered by HowStuffWorks
What is spam-filtering software?
Answered by Discovery Channel
Where did the idea for Coffee & Power come from?
Answered by Philip Rosedale