Storing ID related fields in influxdb

We are planning to use influxdb for our solution, however for our implementation we have some questions . It will help us if you will clarify the same.

1.) Our usecase contains tenant specific data. What is the optimal way of persisting tenantID ?
- Should we use separate measurement for each tenant?
- Can we use TAG for storing tenant-ID?
- Should we prefix TenantID to time series data?
2) As per guidelines we should keep searchable fields as tags. However, we have more than 5 searchable fields and 3 of them are id based values (e.g transaction id) - Is it recommended to define TAG type for all String type searchable fields?

@chintan_vaghani This is an issue many Influx users run into.

  1. I would use tenantID as a Tag. The concern with storing ID as tags is series cardinality.

  2. What do you mean by searchable fields? If the tags are dependent (see series cardinality link for explanation) then adding them is very easy. However if they are not dependent then this can cause issues.

Can you share the approximate # of transationIDs tenantIDs and the other UUIDs you are looking to store?