I don't particularly like the nested entities approach because from what I recall it will execute separate SQL queries for each top level record which, to me, doesn't seem very ideal for large scale indexing.
I know it's a pain to do a ton of joins.. believe me our dataset has a boat load of joins too but I think it works out much better to have a GIANT SQL statement execute because you can do record level fetching and index faster (as opposed to waiting for the entire recordset to buffer and send to the client). Try using temporary tables in your connection to help reduce some of the data down or stored procedures if you have that control in your DB. Hope that helps! Amit On Thu, Oct 29, 2009 at 5:00 PM, Avlesh Singh <avl...@gmail.com> wrote: > Read this example fully - > http://wiki.apache.org/solr/DataImportHandler#Full_Import_Example > "nested entities" is an answer to your question. The example has a sample. > > Cheers > Avlesh > > On Fri, Oct 30, 2009 at 2:58 AM, Joel Nylund <jnyl...@yahoo.com> wrote: > > > Hi, Its been hurting my brain all day to try to build 1 query for my > index > > (joins upon joins upon joins). Is there a way I can do multiple queries > to > > populate the same index? I have one main table that I can join everything > > back via ID, it should be theoretically possible.... > > > > If this can be done, can someone point me to an example? > > > > thanks > > Joel > > > > >
