In one db there are some items that are too large for the index and so they get skipped ... here is a log snippet ...

Tue Mar 30 16:18:03 future-production.system.indexes Btree::insert: key too large to index, skipping future-production.quotations.$graph_1 1453 { : { rdf:

{ type: "quotation:quotation" }

, dcterms:

{ title: null, source: "U.S. Congressional Record" }

, quotation:

{ said_on: new Date(1173373205000), reporter: "http://pius.me", organizationSpokenTo: "http://dbpedia.org/page/United_States_Senate", subject: null, evidence: "http://frwebgate.access.gpo.gov/cgi-bin/getpage.cgi?position=page&page...", venue: "http://dbpedia.org/page/United_States_Senate", primary_source: "U.S. Congressional Record", speaker_bioguide_id: "300047", body: "I think after a lot of give and take and after a lot of court decision..." }

} }

When this happens and you query later on that collection that would use the index what does mongodb do? Does it just omit those documents in the results since they didn't make it into the index ... or does it skip that index and do scans since it knows it has missing documents?