Hi Paul, You're seeing properties fragments. For each document fragment, there is also a fragment to hold metadata properties that you can manipulate with xdmp:document-set-properties()
There should not be any hidden documents. It's possible that the metric under "Documents" is actually counting document-space fragments (as opposed to properties fragments.) That would be a bug. The reason we would not have noticed this recently is that we now say, rather vehemently -- don't use document fragments. Customers have far more success and less confusion when you split the documents up when ingesting. If you need to get the large XML docs back out as is, you may want to write something to reassemble them from smaller documents. Charles ________________________________ From: general-boun...@developer.marklogic.com [general-boun...@developer.marklogic.com] on behalf of Paul Vanderveen [pvanderv...@terraxml.com] Sent: Wednesday, September 07, 2016 11:35 AM To: MarkLogic Developer Discussion Subject: [MarkLogic Dev General] General Question about Documents and Fragments I have a question on what MarkLogic considers to be a document. My database contains many large XML documents, so we are using fragmenting. My database summary is as follows: Documents: 281,000 Fragments: 569,000 [cid:image001.jpg@01D20904.45F9BE80] However, if I do this query: count(/*) I only get 39,000 results. This seems about right based on documents that we have put in. So when MarkLogic reports 281,000 documents, what is it considering as a document? Are there “hidden” documents that are used by MarkLogic for indexing and other purposes? Paul Vanderveen
_______________________________________________ General mailing list General@developer.marklogic.com Manage your subscription at: http://developer.marklogic.com/mailman/listinfo/general
