Database Time Series - Metrics vs. Tags

Question

I am new to TSDB and I have many temperature sensors to store in my database at one point per second. Is it better to use one unique metric per sensor or only one metric (e.g. temperature) with different labels depending on the sensor?

I searched on the Internet what is the best practice, but I did not find a good answer ...

Thank!: -)

Edit: I will have 8 types of measurements (temperature, setpoint, energy, power, ...) from 2500 sources.

Answer 1

InfluxDB, , , . , , InfluxDB .

TSDB, , .

Answer 2

, Axibase Time-Series Database . ATSD , , . , , , .. , , .

, 0601911 - EPA. : http://axibase.com/environmental-monitoring-using-big-data/.

, , , , .. .

1: :

series e:sensor-001 d:2015-08-03T00:00:00Z m:temperature=42.2 m:humidity=72 m:precipitation=44.3

, -001, , .., , . /, , "", /. .

series e:sensor-001 d:2015-08-03T00:00:00Z m:temperature=42.2 ... t:status=Provisional

Answer 3

. , , , (, ).

, , , , , , , . , ( 16 OpenTSDB, ).

Tags distinguish between these pieces of data. For example, you can mark data in different ways if it suddenly changes dramatically, so that if necessary you get only relevant data without losing the rest of the data. Although for a temperature sensor receiving data every second, it would probably be best to filter and store data only when the value has changed ...

Best practices are summarized here.