Question

Let’s presume I have objects stored in MongoDB with the following structure:

Transaction
{
  _id
  userId
  accountId
}

And assume I have the following index:

db.Transaction.ensureIndex({"userId": 1})

Does the following query take any advantage of the index to minimize search time?

db.Transaction.find( {userId: 'user1234', accountId: 'account1234'} );

That is, does MongoDB use the index to “whittle down” the results by userId and then table-scan for accountId?

db.Transaction.find( {userId: 'user1234', accountId: 'account1234'} ).explain()
{
    "cursor" : "BtreeCursor userId_1",
    "nscanned" : 2,
    "nscannedObjects" : 2,
    "n" : 1,
    "millis" : 1,
    "nYields" : 0,
    "nChunkSkips" : 0,
    "isMultiKey" : false,
    "indexOnly" : false,
    "indexBounds" : {
            "userId" : [
                    [
                            "user1234",
                            "user1234"
                    ]
            ]
    }

Looking at the explain() for the query says BtreeCursor userId_1, so I presume it got all the users with userId of user1234 and then scanned (the only two items) to find the accountId of account1234 – Is this correct?

Thanks in advance.

Answer 1

Does the following query take any advantage of the index to minimize search time?

Yes, it does.

Looking at an explain() for the query says BtreeCursor userId_1, so I presume it got all the users with userId of user1234 and then scanned to find the accountId of account1234 – Is this correct?

Yes, you are correct. See here for more information:

• http://www.mongodb.org/display/DOCS/Indexes
• http://www.mongodb.org/display/DOCS/Indexing+Advice+and+FAQ