I agree. Well done! You may also want to look at how Vertica implements parallel analytics on top of a column based store commercially: http://www.vertica.com/_pdf/VerticaArchitectureWhitePaper.pdf
For example, this is interesting: "Logical tables are decomposed and physically stored as overlapping groups of columns, called "projections," and each projection is sorted on a different attribute (or set of attributes), which optimizes them for answering queries with predicates on its sort attributes." Call that insert-time view materialization? Clint Morgan's THBase contrib is maybe the first step down the road to something like that done with an eager strategy. A lazy approach could do something in the background whenever a WAL is rolled. - Andy ________________________________ From: stack <[email protected]> To: [email protected] Sent: Friday, August 7, 2009 8:55:01 AM Subject: Re: HBase Analytics: A New DB for 80% of Facebook, YouTube-scale Sites Its excellent. St.Ack On Fri, Aug 7, 2009 at 8:43 AM, Bradford Stephens < [email protected]> wrote: > Hey all, > > I'd really appreciate your feedback on an article I just wrote on a > new kind of DB, build on HBase, to serve massive-scale websites and do > low-latency analytics. > > http://www.roadtofailure.com/2009/08/07/a-new-db-for-80-of-facebook-youtube-scale-sites/ > > Feel free to leave feedback, even if you vehemently disagree! Greatly > appreciated. > > -- > http://www.roadtofailure.com -- The Fringes of Scalability, Social > Media, and Computer Science >
