I have Influxdb (v 1.7.1) table with schema like this one:
table_name tags: tag_1, tag_2, tag_3 fields: field_1, field_2, field_3
all fields are varying a lot(millions different values) so cannot use them as tags, get into to many series errors
Result i need is count of distinct combinations between
field_2 grouped by days, and some tags
On PostgreSQL we achieve this with grouping
field_1 and then counting distinct
field_2 , rows are numerous and queries are getting to slow.
Im stuck with with distinct combinations between two fields, grouping part is easy. Distinct works only with one field influxdb. Idea was to concat them into another series and then count, influx has no such thing.
field_1 value i have hundred thousands of not unique
field_2 values. I have millions of
field_2 values, so moving them to tags and grouping is out of the question.
I need a strategy of possible aggregation of data till i get to that count.