Hi, I am trying out SolrCloud using a recent Solr 4 nightly. We use result grouping (FieldCollapsing) and found that the value of ngroups returned by Solr is not correct.
My understanding is that all the documents belonging to the same group should be on the same shard to ensure that ngroups returns the correct value. However, the shard that a document is sent to is decided automatically based on the value of the uniqueKey field. Is it possible for Solr to hash fieldX instead of the uniqueKey while distributing the documents to the different shards? Is there some other way of getting accurate values of ngroups when using SolrCloud? Thanks, Nikhil
