mongodb的group和aggregate问题记录

1.group只支持单机分组，不支持多分片集群分组，这是个大坑

2.aggregate支持多分片集群，新版本建议使用aggregate

3.数据模板

{
    "_id": "2204000000429520201204000000000",
    "mp_id": "22040000004295",
    "create_time": ISODate("2020-12-04T16:10:03.627Z"),
    "data_day": NumberInt("20201204"),
    "data_time": ISODate("2021-03-11T01:29:55.095Z"),
    "gd_id": "125F58D3-B632-43B0-B6D8-1EA90977F33C",
    "mn": "20150724000002",
    "protocol": "0",
    "update_time": ISODate("2020-12-04T16:10:03.627Z"),
    "values": [
        {
            "cou_value": null,
            "flag": null,
            "avg_value": null,
            "mstatus": NumberInt("6"),
            "pollutant_code": "001",
            "min_value": null,
            "avg_zs": null,
            "estatus": NumberInt("0"),
            "astatus": NumberInt("0"),
            "outlet_standard": "20",
            "max_zs": null,
            "min_zs": null,
            "max_value": null,
            "sstatus": NumberInt("0")
        },
        {
            "cou_value": null,
            "flag": null,
            "avg_value": null,
            "mstatus": NumberInt("6"),
            "pollutant_code": "002",
            "min_value": null,
            "avg_zs": null,
            "estatus": NumberInt("0"),
            "astatus": NumberInt("0"),
            "outlet_standard": "80",
            "max_zs": null,
            "min_zs": null,
            "max_value": null,
            "sstatus": NumberInt("0")
        },
        {
            "cou_value": null,
            "flag": null,
            "avg_value": null,
            "mstatus": NumberInt("6"),
            "pollutant_code": "003",
            "min_value": null,
            "avg_zs": null,
            "estatus": NumberInt("0"),
            "astatus": NumberInt("0"),
            "outlet_standard": "250",
            "max_zs": null,
            "min_zs": null,
            "max_value": null,
            "sstatus": NumberInt("0")
        },
        {
            "cou_value": null,
            "flag": null,
            "avg_value": null,
            "mstatus": NumberInt("6"),
            "pollutant_code": "004",
            "min_value": null,
            "avg_zs": null,
            "estatus": NumberInt("0"),
            "astatus": NumberInt("0"),
            "outlet_standard": "80",
            "max_zs": null,
            "min_zs": null,
            "max_value": null,
            "sstatus": NumberInt("0")
        },
        {
            "cou_value": null,
            "flag": null,
            "avg_value": null,
            "mstatus": NumberInt("6"),
            "pollutant_code": "008",
            "min_value": null,
            "avg_zs": null,
            "estatus": NumberInt("0"),
            "astatus": NumberInt("0"),
            "outlet_standard": "50",
            "max_zs": null,
            "min_zs": null,
            "max_value": null,
            "sstatus": NumberInt("0")
        },
        {
            "pollutant_code": "B02",
            "cou_value": null,
            "min_value": null,
            "flag": null,
            "avg_value": null,
            "outlet_standard": null,
            "max_value": null
        }
    ]
}

4.group分组(cond筛选条件,key是分组的字段，1标示true按此字段分组, initial初始化字段值，reduce对分组数据进行循环处理，item代表分组的每行数据，out包含要输出的字段,item.values.length代表子集的总数)；

var start_time = ISODate("2021-03-10T16:05:46.186+0800");
var end_time = ISODate("2021-03-30T16:06:01.666+0800");
 
db.collection.group({
    cond: {
        "$and": [
            {
                "data_time": {
                    "$gte": start_time
                }
            },
            {
                "data_time": {
                    "$lte": end_time
                }
            }
        ]
    },
    key: {
        "mp_id": 1,
        "total": 1,
        "type": 1
    },
    initial: {
        "total": 0,
        "type": "day_air"
    },
    reduce: function Reduce(item, out) {
        if (item.values != null) {
            out.total += item.values.length
        }
    }
});

查询结果如下

[
	{
		"mp_id" : "1101050031132015",
		"total" : 64,
		"type" : "day_air"
	},
	{
		"mp_id" : "13810000000645",
		"total" : 208,
		"type" : "day_air"
	},
	{
		"mp_id" : "13810000000665",
		"total" : 208,
		"type" : "day_air"
	},
	{
		"mp_id" : "138200000001115",
		"total" : 208,
		"type" : "day_air"
	},
	{
		"mp_id" : "14010600000115",
		"total" : 208,
		"type" : "day_air"
	}
]

5.aggregate管道聚合方式($match筛选条件,group为分组条件,id必选，可以为空或者Null,total则是自定义聚合字段，我这取的子集的size，$values代表子集，$addFields可以添加自定义的字段，out:输出的指定文档)。

var start_time = ISODate("2021-03-10T16:05:46.186+0800");
var end_time = ISODate("2021-03-30T16:06:01.666+0800");
 
db.collection.aggregate([{ $match : {"$and": [
            {
                "data_time": {
                    "$gte": start_time
                }
            },
            {
                "data_time": {
                    "$lte": end_time
                }
            }
        ]  }},
    {
        $group: {
            _id: "$mp_id",
            total: {
                $sum: {
                    $size: "$values"
                }
            }
        }
    },
    {
        $addFields: {
            "type": "day_air"
        }
    },{
			$out:"temp"
		}
]);

输出结果为