> Is this correct or there are configuration for pushdown mechanism for Solr?
Currently no. The only aggregation that the solr translator currently understands is count star.
> Also there is a limit on the number of row returned. for example if the total number of records is 16000, Solr is quered mutliple times each time returned 1024, like paging.. how to stop that ?
It looks like from the source if a limit is not used, then a batching mechanism of SolrQuery will be used to process the query. If this is causing unintended overhead, then the strategy should be updated.
Thanks for your reply.
I am using count(*) but grouping is not passed to solr.
select x,y, count(*) from solr_core group by x,y;
The Solr log does not show that the group flag is set to true although here count(*) is used.
Sorry I should have been explicit. The group by clause is currently not supported. So you can only issue a count against the entire row set - select count\(*) from solr_core
Can you capture your needs in an enhancement request - JBoss Issue Tracker