BigTable
From Just Solve the File Format Problem
(Difference between revisions)
Dan Tobias (Talk | contribs) (Created page with "{{FormatInfo | formattype=electronic | subcat=Databases | released=2004 }} '''BigTable''' is a database system used by Google in the back end of various applications. It is bu...") |
Dan Tobias (Talk | contribs) |
||
| Line 4: | Line 4: | ||
| released=2004 | | released=2004 | ||
}} | }} | ||
| − | '''BigTable''' is a database system used by Google in the back end of various applications. It is built on the [[Google | + | '''BigTable''' is a database system used by Google in the back end of various applications. It is built on the [[Google File System]]. Neither BigTable nor GFS are available directly to the public, but they are integral to various Google services. |
BigTable is not a relational database system, but has a three-dimensional structure with two arbitrary strings as row and column indices, and a third dimension by timestamp. These coordinates index arbitrary arrays of bytes in a structure designed to scale up to petabyte levels. The data is split into "tablets" of approximately 200 megabytes each, which can be spread across many machines on a network. | BigTable is not a relational database system, but has a three-dimensional structure with two arbitrary strings as row and column indices, and a third dimension by timestamp. These coordinates index arbitrary arrays of bytes in a structure designed to scale up to petabyte levels. The data is split into "tablets" of approximately 200 megabytes each, which can be spread across many machines on a network. | ||
Latest revision as of 02:26, 26 April 2014
BigTable is a database system used by Google in the back end of various applications. It is built on the Google File System. Neither BigTable nor GFS are available directly to the public, but they are integral to various Google services.
BigTable is not a relational database system, but has a three-dimensional structure with two arbitrary strings as row and column indices, and a third dimension by timestamp. These coordinates index arbitrary arrays of bytes in a structure designed to scale up to petabyte levels. The data is split into "tablets" of approximately 200 megabytes each, which can be spread across many machines on a network.
