I'm not sure if this has been addressed or not ...
When you querying for a set of JSON documents, you can select a portion of the doc, or the entire doc. You can also apply filters to the doc so that only documents that match the filters are returned.
Do the filters get applied on the cluster (server side) or on the client prior to the client app filtering the result set?
I want to say that the JSON document APIs do the filtering on the server before returning the doc(s) in a result set, but I need to confirm this.
At the same time... if dealing with a very large document, where does the filtering on the subset of a document occur?
An example would be if I want to find just the company name and addresses of all orders placed within the last 6 months and withing a specific state.
Do I get the entire doc sent back to the client and then the results are filtered to return only the client app?
Note: If MapRDB behaves like HBase, the concept of client/server is a bit different from what we think about when dealing with traditional databases.