Hi,
If Tag is not directly distinct from MeasurementMNode we will need to do an extra check to determine ownership, and it will be harder to check if the path is valid. Therefore, I suggest that the design of the statement can be like this: select * from root.sg1.d1 where tag( tag1 = ??1' and/or tag2 = ??2') and/or s1 > 100 Dawei Liu On 06/10/2021 14:08??HW-Chao Wang<[email protected]> wrote?? Your idea is good. We can provide a query statement to find the data according to the tag. For example, select * from root. A.B where tag1 = 10 ------------------ ???????? ------------------ ??????: "dev" <[email protected]>; ????????: 2021??6??9??(??????) ????8:12 ??????: "[email protected]"<[email protected]>; ????: how about support query data by tag Hi, At present, We support using tags and attributes to add additional attributes to timeseries, but in fact, tags and attributes are quite similar to influxdb's tags and fields. Suppose that when I create timeseries, I add additional tags and attributes to uniquely identify the timeseries. When querying, you can find the timeseries according to tags and attributes, and then query the data according to the timeseries. Can we provide a syntax to query data directly according to tags or attributes? Thanks! Chao Wang BONC ltd [email protected]
