Hi, I have a Solr index with an awful lot of tiny documents in it (hundreds of millions) each having a few fields.
I can query the index without problem and retrieve results sorted on score. I am now exploring the possibility of sorting on other fields. The typical number of results from queries is in the 100s or 1000s, so tiny in regard to index size. >From what I remember of Lucene, it had (still has?) to build a FieldCache prior to sorting, and this FieldCache contains all field values for the field to sort on right? My question is, is there a way around that behaviour of the field cache, i.e. could a filter query somehow limit the values loaded in the FieldCache? Is there otherwise a WTH (Well Known Hack) to be able to sort on fields when an index has lots (100s of 1000s) of values for that field? Any help appreciated, Mathias.
