我的收藏的Azure Cosmos DB索引策略 [英] Azure Cosmos DB indexing policy for my collection

问题描述

我是Cosmos DB的新手，我正在将传感器数据推送到IoT中心，然后再推送到Cosmos DB.

Hi I am new to Cosmos DB , I am pushing my sensor data to IoT hub and then to Cosmos DB .

以下哪种格式最好将其存储在宇宙中以增加读取和写入的吞吐量，因为在每条记录中，我将具有150个传感器和

which of the following format will be better one to storing the data in cosmos to increase the throughputs of reads and writes as in each record i will be having 150 number of sensors and 

几乎所有我对DeviceId的查询(这也是分区键).和currentTime

And Almost all my queries on DeviceId ( which is partition key also ) . and currentTime

{
"DeviceId" : 1234 , 
"sensor1" : 234,
"sensor3" : 4,
....
....
"currentTime" : 121212
}

{
  "DeviceId" : "asad",
  "currentTime" : 1212123,
  "data" : {
        "sensor1" : 323,
        "sensor2" : 234,
        "sensor3" : 444,  
       ......
       ......
   }
}

据我所知，如果我使用第一个，它将索引每个键，并且读写操作会很昂贵，因为 甚至我将索引Path都保留为/，因为所有索引都处于同一级别，每个键都将是索引.

As I know if I use first one it will index each key and reads and writes will be costly , as    even I keep include index Path as / , as all are at same level each key will be index . 

如果我使用第二个，我将只能控制仅对一个索引编制索引并排除对/data的索引编制?

If i use second one i will have control over to index only only one level and Exclude indexing of /data?

任何建议，因为我对cosmos DB相对较新.

Any suggestion as i am relatively new to cosmos DB.

推荐答案

您好，Amjathkhan，

Hi Amjathkhan,

请在上阅读 索引路径.我将在您的第二个Json示例中使用类似下面的内容，但是取决于如果您的Json和分区集合中有数组等. 

Please read-up on Index Paths. I would use something like the below with your second Json example but, depends  if you have arrays in your Json and partitioned collections, etc. 

/prop/*

Index path for all paths under the specified label. Works with the following queries

SELECT FROM collection c WHERE c.prop = "value"
SELECT FROM collection c WHERE c.prop.subprop > 5
SELECT FROM collection c WHERE c.prop.subprop.nextprop = "value"
SELECT FROM collection c ORDER BY c.prop

/props/[]/subprop/?

/props/[]/subprop/?

Index path required to serve iteration and JOIN queries against arrays of objects like [{subprop: "a"}, {subprop: "b"}]:

SELECT tag FROM tag IN collection.props WHERE tag.subprop = "value"
SELECT tag FROM collection c JOIN tag IN c.props WHERE tag.subprop = "value"

这篇关于我的收藏的Azure Cosmos DB索引策略的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！