Question/advice how to setup influxdb tags/fields

InfluxDB 2
,
1

I have a question about the usage ot tags / fields. I know that Tags are indexed

So i ám collectiong data from approx 120 measurement now
All buidings are in tags Area, al parts in that buildings have a second tag Part.
Now some processes are batch, each process geta q unique ID number
I have 2 solutions
1 Store the batchnumber as tagvalue or as fieldvalue
batch duration for short ones about 24hrs, others are 5 days

To view the data later back in Grafana i want to select the batchnumber from a variable query to display the corresponding data in a graph.

Bucket ActualValeus retention 6 months
Area= | Part=
BuildingA | PartA
BuildingA | PartB
BuildingA | PartC
BuildingB | PartX
BuildingB | Part Y

Bucket Counters retention 5 years
Area= | Part=
BuildingA | PartA
BuildingA | PartB
BuildingA | PartC
BuildingB | PartX
BuildingB | Part Y

2

Hello @Admini,
Sorry what exactly is your question?
If your unique ID number is unbounded, then you should definitely store that as a field to avoid runaway cardinality problems. Unless of course you’re entirely expiring/deleting the data after short periods, then it might be okay to store it as a tag. If you’re simply downsampling, then you’ll still have a bunch of tags and your cardinality will grow exponentially.
Influxdb instance size is not directly related to cardinality. You can have a cardinality of 3 and write a million points per second. Or you can have a cardinality of 1000 and be writing 1000 points per second.

3

Sorry for the delay, we used a historian to collect data for this. A historian knows no tags, only fields.
I make the batch number as a field and put the new batch number here every day