Tech Brief: NoSQL for DBAs

NoSQL describes several database technologies for which the data stores don&#039;t need a fixed schema and often avoid join operations. Brian Aker gives a lightning talk (7 minutes) about NoSQL during the Nov 2009 OpenSQLCamp in Portland, Oregon.<br><br><embed src=&#034;http://www.youtube.com/v/LhnGarRsKnA&hl=en_US&fs=1&#034; type=&#034;application/x-shockwave-flash&#034; width=&#034;425&#034; height=&#034;344&#034; allowscriptaccess=&#034;always&#034; allowfullscreen=&#034;true&#034;></embed><br><br>Data operations and manipulation are often performed via map/reduce operations, which would most closely map to &#034;group by&#034; operations in a traditional RDBMS. In general, Brian talks about how NoSQL data operations are best implemented in computational networks by dividing the operations among multiple machines.<br><br>Perhaps one of the hardest concepts to grasp by people with traditional database backgrounds is that NoSQL treats data in its natural order as optimal and applies transformations to it. This runs contrary to the RDBMS view of the world.<br><br>The biggest take away from the presentation is that NoSQL databases are ideal for situations where system requirements define non-transactional data manipulation -- a quality of many web-based systems.<br><br>The recording quality is not great in some portions, where comments from the audience mask the audio from the speaker.<br><br>The <a href=&#039;http://www.opensqlcamp.org/&#039; target=&#039;_blank&#039;>notes</a> and the <a href=&#039;http://slidesha.re/d5GQv&#039; target=&#039;_blank&#039;>slides</a> for the presentation are available for review.").substring(0,300); return _diggURL ; } NoSQL describes several database technologies for which the data stores don't need a fixed schema and often avoid join operations. Brian Aker gives a lightning talk (7 minutes) about NoSQL during the Nov 2009 OpenSQLCamp in Portland, Oregon.

Data operations and manipulation are often performed via map/reduce operations, which would most closely map to "group by" operations in a traditional RDBMS. In general, Brian talks about how NoSQL data operations are best implemented in computational networks by dividing the operations among multiple machines.

Perhaps one of the hardest concepts to grasp by people with traditional database backgrounds is that NoSQL treats data in its natural order as optimal and applies transformations to it. This runs contrary to the RDBMS view of the world.

The biggest take away from the presentation is that NoSQL databases are ideal for situations where system requirements define non-transactional data manipulation -- a quality of many web-based systems.

The recording quality is not great in some portions, where

Click Here to read more...