Space use when dealing with tags


I got a dilemma over dealing with tags.

Let’s say I write a huge number points into same measurement. How does the use of tags affect space use and performance: all points are stored without tag VERSUS all points are stored under a singe tag (= 100% high cardinality)?

Currently we put tag just under first point written and the following are without tags but that is due to optimization of API payload.



Hello @sampler,
Great question. I recommend checking out this documentation about dependent tags and how the use of tags can decrease cardinality.


Thank you for reply but I’m afraid I can’t find exact answer to my question.

I have a data series of 100000 values and timestamps. I have two options:
a) insert data and add a constant tag to each point (same tag for all points)
b) insert data without the tag

Cardinality is ideal but i doubt the footprint is same in both cases?