We have used both Solr and graph databases for our XML file indexing. Both are equivalent in terms of performance, but a graph db (such as Neo4j) offers a lot more flexibility in joining across the nodes and traversing. If your data is strictly hierarchical Solr might do it, alternately suggest looking at a Graph database such as Neo4j.
On Fri, Apr 27, 2012 at 10:36 AM, Bob Sandiford < bob.sandif...@sirsidynix.com> wrote: > ndexing and searching of the specific fields, it is certainly possible to > retrieve the xml file. While Solr isn't a DB, it does allow a binary field > to be associated with an index document. We store a GZipped XML file in a > binary field and retrieve that under certain conditions to get at original > document information. We've found that Solr can handle these much faster > than our DB can do. (We regularly index a large portion of our documents, > and the XML files are prone to frequent changes). If you DO keep such a > blob in your Solr index, make sure you retrieve that field ONLY when you > really want it...
