Dedupe records in Mongo with mongoose

findDupes.js

db.downloads.aggregate(
    { '$group': { _id: '$userId', total: { '$sum': 1 } } },
    { '$match': { total: { '$gte': 2 } } }
);

mergeDupes.js

_.forEach(dupeIds, function(dupeId) {
    db.find({ userId: dupeId }, function(err, result) {
        if(err || !result || result && result.length < 2) {
            console.log('skipping ', dupe);
        } else {
            //manually dedupe into first record, this is custom to your schema
            var resources1 = _.indexBy(result[0].resources, 'resourceId');
            var resources2 = _.indexBy(result[1].resources, 'resourceId');
            var mergedResources = _.values(_.merge(resources1, resources2));
            result[0].resources = mergedResources;
            
            //save first, remove duplicate
            result[0].save(function(err) {
            if(err) console.error(err);
                result[1].remove(function(err) {
                    if(err) console.error(err);
                });
            });
        }
    });
});