Hi, I never used map-reduce indexing.
My understanding is that map-reduce tasks generate one or more Solr indices, then the golive tool is used in order to merge these indices at core level to one or more shards (the shard's leaders) in a Solrcloud collection. After merge occurs in leaders the replica are automatically synchronized. My question is what about document routing ? If I want to route documents to shards according to one field of my documents, how can this work if the merge is done at core level ? Thank you. Dominique
