Comments on: Column stores http://brunodumon.wordpress.com/2007/10/07/column-stores/ Bruno's blog about the Daisy CMS and the Kauri webapp framework Sun, 29 Mar 2009 18:28:37 +0000 http://wordpress.com/ hourly 1 By: Raymond Brown http://brunodumon.wordpress.com/2007/10/07/column-stores/#comment-34 Raymond Brown Sun, 04 Nov 2007 11:14:52 +0000 http://brunodumon.wordpress.com/2007/10/07/column-stores/#comment-34 Your post set prompted me to look into several solutions. I ended up surveying the bulk of the xml database field. The end result that I wanted to pass by you, is, why not use IBM's DB2 Express-C 9.5 (a free edition, a mature platform on various OSs) and transition the Daisy repository to a pure xml db? DB2 would allow a transition path from the relational model, using the current Daisy relational db schema, to a pure xml-based repository solution, in the same database. See page 55, Fig 2.2 of "http://www.redbooks.ibm.com/abstracts/sg247301.html?Open" for where I'm coming from, along with the surrounding explanation. The end solution would allow managing entities with a flexible set of attributes (Daisy documents) using a non-relational, XML databaase with Daisy Query mapped to XQuery with its performance attributes. This has been fun. Daisy is great! Your post set prompted me to look into several solutions. I ended up surveying the bulk of the xml database field. The end result that I wanted to pass by you, is, why not use IBM’s DB2 Express-C 9.5 (a free edition, a mature platform on various OSs) and transition the Daisy repository to a pure xml db? DB2 would allow a transition path from the relational model, using the current Daisy relational db schema, to a pure xml-based repository solution, in the same database.

See page 55, Fig 2.2 of “http://www.redbooks.ibm.com/abstracts/sg247301.html?Open” for where I’m coming from, along with the surrounding explanation.

The end solution would allow managing entities with a flexible set of attributes (Daisy documents) using a non-relational, XML databaase with Daisy Query mapped to XQuery with its performance attributes.

This has been fun. Daisy is great!

]]> By: John Sichi http://brunodumon.wordpress.com/2007/10/07/column-stores/#comment-28 John Sichi Sun, 28 Oct 2007 17:15:57 +0000 http://brunodumon.wordpress.com/2007/10/07/column-stores/#comment-28 Check out LucidDB also. Check out LucidDB also.

]]> By: brunodumon http://brunodumon.wordpress.com/2007/10/07/column-stores/#comment-10 brunodumon Tue, 09 Oct 2007 17:31:10 +0000 http://brunodumon.wordpress.com/2007/10/07/column-stores/#comment-10 Thanks for the links. BTW, currently these are just some random musings, no actions planned. Thanks for the links.

BTW, currently these are just some random musings, no actions planned.

]]> By: Ross http://brunodumon.wordpress.com/2007/10/07/column-stores/#comment-9 Ross Tue, 09 Oct 2007 14:39:01 +0000 http://brunodumon.wordpress.com/2007/10/07/column-stores/#comment-9 Given Daisy's document-centric approach, something like <a href="http://www.couchdbwiki.com/index.php?title=Main_Page" rel="nofollow">CouchDB</a> might also be appropriate. Although it's probably not quite ready for "prime time" yet, it might be worthwhile to continue watching this space. I'm <em>very</em> interested to see where you go with this. Given Daisy’s document-centric approach, something like CouchDB might also be appropriate. Although it’s probably not quite ready for “prime time” yet, it might be worthwhile to continue watching this space.

I’m very interested to see where you go with this.

]]> By: ian http://brunodumon.wordpress.com/2007/10/07/column-stores/#comment-8 ian Mon, 08 Oct 2007 08:31:05 +0000 http://brunodumon.wordpress.com/2007/10/07/column-stores/#comment-8 Interesting topic! I've often thought about Daisy's storage model in this context too. You may also be interested in the column store work at the Hadoop project, which includes an RDF api through Hbase. http://wiki.apache.org/lucene-hadoop/Hbase This recent write up on Amazon's Dynamo project may be of interest too: http://allthingsdistributed.com/2007/10/amazons_dynamo.html Interesting topic! I’ve often thought about Daisy’s storage model in this context too. You may also be interested in the column store work at the Hadoop project, which includes an RDF api through Hbase.
http://wiki.apache.org/lucene-hadoop/Hbase

This recent write up on Amazon’s Dynamo project may be of interest too: http://allthingsdistributed.com/2007/10/amazons_dynamo.html

]]>