In a word, no. At least probably not. There are some JIRA tickets dealing with distributed joins, and some with certain restrictions, specifically if the second (from) collection can be reproduced on every slice of the first (to) collection.
In the trunk (6.0), there's the ParallelSQL stuff which has some relevance, but it's still not a full RDBMS type join. The usual recommendation is to flatten your data if at all possible so you don't _have_ two collections. Solr is a wonderful search engine. It is not an RDBMS and whenever I find myself trying to make it behave like an RDBMS I try to rethink the architecture. On Mon, Nov 16, 2015 at 6:56 PM, soledede_w...@ehsy.com <soledede_w...@ehsy.com> wrote: > Dear @solr_lucene > currently,I am using solr5.3.1,I have a requirement, I need search like in > relation database(select * from A ,B where A.id=B.id),Can we implments with > solr5.3 in SolrCloud mode,I have two collection,2 shards per collection. > Help me please. > > Thanks > > > soledede_w...@ehsy.com
