In mongodb, some values ​​in mapReduce are not processed by reduce.

Question

When using mapReduce in mongodb, some values ​​are not processed by reduce, which is very confusing. First code

    db.test.mapReduce(
        function () {
            emit({
                host: this.host,
                os: this.os,
                computed_on_date: this.computed_on_date,
                uid: this.uid
            }, {data: this.os});
        },
        function (key, value) {
            return {result: 1};
        },
        {
            out: "a"
        }
    )

The data that should be output is

        "_id" : {
                "host" : "www.ddc.com",
                "os" : "android",
                "computed_on_date" : "2017-04-19",
                "uid" : "0ae71c7\ne-7da6-9051-ec6c-49dc671b4e3a"
        },
        "value" : {
                "result" : 1
        }

The actual output data is

      "_id" : {
                "host" : "www.ddc.com",
                "os" : "android",
                "computed_on_date" : "2017-04-19",
                "uid" : "0ae71c7\ne-7da6-9051-ec6c-49dc671b4e3a"
        },
        "value" : {
                "data" : "android"
        }

I am very confused. When the key of condition emit is less than 4, it will be displayed normally. If it is more than 3 or query condition is added, reduce processing will not be performed.

Another question is, when mapReduce has a query, should it process the query first or process reduce and then process the query in the result

Answer 1

It is recommended to use aggregation. Compared with MR, aggregation is the first class feature of MongoDB.

For reference.

Love MongoDB! Have fun!