As in version 1.4, the series concept is: ( link )
In InfluxDB, a series is the collection of data that share a retention policy, measurement, and tag set.
So in the sample data, the series is four.
But from version 1.5 to 1.8, the concept changed, it said series also include field key: ( link )
In InfluxDB, a series is a collection of points that share a measurement, tag set, and field key.
So in the same sample data, the series is eight
I think timestamp、retention policy、measurement and tag set is unique identifier.
e.g:
> INSERT cpu,host=AA,region=BB value1=0.1,value2=0.2 1615431200
> select * from cpu
name: cpu
time host region value1 value2
---- ---- ------ ------ ------
1615431200 AA BB 0.1 0.2
> INSERT cpu,host=AA,region=BB value3=100 1615431200
> select * from cpu
name: cpu
time host region value1 value2 value3
---- ---- ------ ------ ------ ------
1615431200 AA BB 0.1 0.2 100
Hello @tankywoo,
Can you help me understand what problem you are trying to solve? Getting a feel for the bigger picture might help me to answer your question better. Are you trying to calculate series cardinality?
What do you mean by the series is 4 vs the series is 8?
You have a series cardinality of 1 for a series that contains the following unique elements: 1 measurement (cpu), two unique tag key/value pairs (host=AA and region==BB; 11=1 and 11=1), and 3 field keys (value1, value2, and value3).
The example here is useful to me: InfluxDB glossary | InfluxDB OSS 1.8 Documentation
I’m confused with the concept of series, and what I list above, the concept of influxdb v1.4 vs v1.8 is different, you can open the link with the part of series, in old version, the definition of series is controlled by rp, measurement and tag, but in new version, the definition remove rp, but add field key, I use influxdb from 0.9, but I read the new version document recently, the changed definition make me confused, I think the the definition of new version in the document is not right.
