Membase and CouchDB Buckets
From what I’ve been reading membase stores data in vbuckets and these are/can be spread over a number of servers. So each server has a sub-set of the data. However, CouchDB stores everything in a single database which is not spread over any servers (? Is this right ?) which means the ‘view’ mechanism of CouchDB is performed on a server by server bases. Is it proposed that each couchbase instance in a cluster (from the membase world) will have its own CouchDB instance for storing its own partition (vbucket) of the data set? Does this allow for distributed views and distributed map/reduce functions etc?